ai-document-parser 0.1.101 → 0.1.104
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +3 -2
- package/dist/convert/blocks.d.ts +14 -0
- package/dist/convert/blocks.d.ts.map +1 -0
- package/dist/convert/blocks.js +254 -0
- package/dist/convert/blocks.js.map +1 -0
- package/dist/convert/converter.d.ts.map +1 -1
- package/dist/convert/converter.js +84 -62
- package/dist/convert/converter.js.map +1 -1
- package/dist/convert/emit.d.ts +1 -0
- package/dist/convert/emit.d.ts.map +1 -1
- package/dist/convert/emit.js +32 -33
- package/dist/convert/emit.js.map +1 -1
- package/dist/edit/docx-ooxml.d.ts.map +1 -1
- package/dist/edit/docx-ooxml.js +5 -33
- package/dist/edit/docx-ooxml.js.map +1 -1
- package/dist/index.d.ts +1 -1
- package/dist/index.d.ts.map +1 -1
- package/dist/parse/docx-blocks.d.ts +18 -0
- package/dist/parse/docx-blocks.d.ts.map +1 -0
- package/dist/parse/docx-blocks.js +654 -0
- package/dist/parse/docx-blocks.js.map +1 -0
- package/dist/parse/html-blocks.d.ts +14 -0
- package/dist/parse/html-blocks.d.ts.map +1 -0
- package/dist/parse/html-blocks.js +244 -0
- package/dist/parse/html-blocks.js.map +1 -0
- package/dist/parse/pptx.d.ts +0 -1
- package/dist/parse/pptx.d.ts.map +1 -1
- package/dist/parse/pptx.js +95 -24
- package/dist/parse/pptx.js.map +1 -1
- package/dist/parse/router.d.ts.map +1 -1
- package/dist/parse/router.js +25 -6
- package/dist/parse/router.js.map +1 -1
- package/dist/types.d.ts +37 -0
- package/dist/types.d.ts.map +1 -1
- package/package.json +2 -2
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"docx-blocks.js","sourceRoot":"","sources":["../../src/parse/docx-blocks.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,aAAa,EAAE,MAAM,oBAAoB,CAAC;AAEnD,OAAO,EAAE,yBAAyB,EAAE,MAAM,sBAAsB,CAAC;AAqBjE,SAAS,iBAAiB,CAAC,CAAS;IAClC,OAAO,CAAC;SACL,OAAO,CAAC,OAAO,EAAE,GAAG,CAAC;SACrB,OAAO,CAAC,OAAO,EAAE,GAAG,CAAC;SACrB,OAAO,CAAC,SAAS,EAAE,GAAG,CAAC;SACvB,OAAO,CAAC,SAAS,EAAE,GAAG,CAAC;SACvB,OAAO,CAAC,QAAQ,EAAE,GAAG,CAAC,CAAC;AAC5B,CAAC;AAED,SAAS,IAAI,CAAC,EAAU,EAAE,IAAY;IACpC,MAAM,EAAE,GAAG,IAAI,MAAM,CAAC,GAAG,IAAI,YAAY,EAAE,GAAG,CAAC,CAAC;IAChD,MAAM,CAAC,GAAG,EAAE,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACtB,IAAI,CAAC;QAAE,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC;IACnB,MAAM,GAAG,GAAG,IAAI,MAAM,CAAC,GAAG,IAAI,YAAY,EAAE,GAAG,CAAC,CAAC;IACjD,OAAO,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC;AAC3B,CAAC;AAED,SAAS,gBAAgB,CAAC,GAAW;IACnC,MAAM,CAAC,GAAG,mCAAmC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;IACxD,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC;AACxB,CAAC;AAED,+FAA+F;AAC/F,SAAS,eAAe,CAAC,GAAW,EAAE,KAAa,EAAE,SAAiB;IACpE,MAAM,MAAM,GAAG,IAAI,MAAM,CAAC,MAAM,SAAS,KAAK,EAAE,GAAG,CAAC,CAAC;IACrD,MAAM,OAAO,GAAG,IAAI,MAAM,CAAC,OAAO,SAAS,GAAG,EAAE,GAAG,CAAC,CAAC;IACrD,MAAM,CAAC,SAAS,GAAG,KAAK,CAAC;IACzB,MAAM,KAAK,GAAG,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;IAC/B,IAAI,CAAC,KAAK,IAAI,KAAK,CAAC,KAAK,KAAK,KAAK;QAAE,OAAO,CAAC,CAAC,CAAC;IAE/C,yCAAyC;IACzC,MAAM,EAAE,GAAG,GAAG,CAAC,OAAO,CAAC,GAAG,EAAE,KAAK,CAAC,KAAK,CAAC,CAAC;IACzC,IAAI,EAAE,GAAG,CAAC;QAAE,OAAO,CAAC,CAAC,CAAC;IACtB,IAAI,GAAG,CAAC,MAAM,CAAC,EAAE,GAAG,CAAC,CAAC,KAAK,GAAG;QAAE,OAAO,EAAE,GAAG,CAAC,CAAC;IAE9C,IAAI,KAAK,GAAG,CAAC,CAAC;IACd,IAAI,CAAC,GAAG,EAAE,GAAG,CAAC,CAAC;IACf,OAAO,KAAK,GAAG,CAAC,IAAI,CAAC,GAAG,GAAG,CAAC,MAAM,EAAE,CAAC;QACnC,MAAM,CAAC,SAAS,GAAG,CAAC,CAAC;QACrB,OAAO,CAAC,SAAS,GAAG,CAAC,CAAC;QACtB,MAAM,QAAQ,GAAG,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QAClC,MAAM,SAAS,GAAG,OAAO,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QACpC,IAAI,CAAC,SAAS;YAAE,OAAO,CAAC,CAAC,CAAC;QAC1B,IAAI,QAAQ,IAAI,QAAQ,CAAC,KAAK,GAAG,SAAS,CAAC,KAAK,EAAE,CAAC;YACjD,MAAM,MAAM,GAAG,GAAG,CAAC,OAAO,CAAC,GAAG,EAAE,QAAQ,CAAC,KAAK,CAAC,CAAC;YAChD,IAAI,MAAM,GAAG,CAAC;gBAAE,OAAO,CAAC,CAAC,CAAC;YAC1B,+CAA+C;YAC/C,IAAI,GAAG,CAAC,MAAM,CAAC,MAAM,GAAG,CAAC,CAAC,KAAK,GAAG,EAAE,CAAC;gBACnC,CAAC,GAAG,MAAM,GAAG,CAAC,CAAC;gBACf,SAAS;YACX,CAAC;YACD,KAAK,EAAE,CAAC;YACR,CAAC,GAAG,MAAM,GAAG,CAAC,CAAC;QACjB,CAAC;aAAM,CAAC;YACN,KAAK,EAAE,CAAC;YACR,CAAC,GAAG,SAAS,CAAC,KAAK,GAAG,SAAS,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC;YAC1C,IAAI,KAAK,KAAK,CAAC;gBAAE,OAAO,CAAC,CAAC;QAC5B,CAAC;IACH,CAAC;IACD,OAAO,CAAC,CAAC,CAAC;AACZ,CAAC;AAED,SAAS,aAAa,CAAC,IAAY;IACjC,MAAM,KAAK,GAA8C,EAAE,CAAC;IAC5D,IAAI,CAAC,GAAG,CAAC,CAAC;IACV,OAAO,CAAC,GAAG,IAAI,CAAC,MAAM,EAAE,CAAC;QACvB,MAAM,KAAK,GAAG,IAAI,CAAC,OAAO,CAAC,QAAQ,EAAE,CAAC,CAAC,CAAC;QACxC,MAAM,GAAG,GAAG,IAAI,CAAC,OAAO,CAAC,MAAM,EAAE,CAAC,CAAC,CAAC;QACpC,IAAI,KAAK,GAAG,CAAC,IAAI,GAAG,GAAG,CAAC;YAAE,MAAM;QAChC,IAAI,IAAiB,CAAC;QACtB,IAAI,KAAa,CAAC;QAClB,IAAI,KAAK,IAAI,CAAC,IAAI,CAAC,GAAG,GAAG,CAAC,IAAI,KAAK,GAAG,GAAG,CAAC,EAAE,CAAC;YAC3C,qEAAqE;YACrE,MAAM,KAAK,GAAG,IAAI,CAAC,MAAM,CAAC,KAAK,GAAG,CAAC,CAAC,CAAC;YACrC,IAAI,KAAK,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC;gBACnC,CAAC,GAAG,KAAK,GAAG,CAAC,CAAC;gBACd,SAAS;YACX,CAAC;YACD,IAAI,GAAG,KAAK,CAAC;YACb,KAAK,GAAG,KAAK,CAAC;QAChB,CAAC;aAAM,CAAC;YACN,MAAM,KAAK,GAAG,IAAI,CAAC,MAAM,CAAC,GAAG,GAAG,CAAC,CAAC,CAAC;YACnC,IAAI,KAAK,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC;gBACnC,CAAC,GAAG,GAAG,GAAG,CAAC,CAAC;gBACZ,SAAS;YACX,CAAC;YACD,IAAI,GAAG,GAAG,CAAC;YACX,KAAK,GAAG,GAAG,CAAC;QACd,CAAC;QACD,MAAM,GAAG,GAAG,eAAe,CAAC,IAAI,EAAE,KAAK,EAAE,IAAI,CAAC,CAAC;QAC/C,IAAI,GAAG,GAAG,CAAC,EAAE,CAAC;YACZ,mEAAmE;YACnE,CAAC,GAAG,KAAK,GAAG,CAAC,CAAC;YACd,SAAS;QACX,CAAC;QACD,KAAK,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,GAAG,EAAE,IAAI,CAAC,KAAK,CAAC,KAAK,EAAE,GAAG,CAAC,EAAE,CAAC,CAAC;QAClD,CAAC,GAAG,GAAG,CAAC;IACV,CAAC;IACD,OAAO,KAAK,CAAC;AACf,CAAC;AAED,SAAS,SAAS,CAAC,OAA2B;IAC5C,MAAM,GAAG,GAAG,IAAI,GAAG,EAAkB,CAAC;IACtC,IAAI,CAAC,OAAO;QAAE,OAAO,GAAG,CAAC;IACzB,MAAM,EAAE,GAAG,wBAAwB,CAAC;IACpC,IAAI,CAAyB,CAAC;IAC9B,OAAO,CAAC,CAAC,GAAG,EAAE,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACvC,MAAM,EAAE,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;QAChB,MAAM,EAAE,GAAG,IAAI,CAAC,EAAE,EAAE,IAAI,CAAC,CAAC;QAC1B,MAAM,MAAM,GAAG,IAAI,CAAC,EAAE,EAAE,QAAQ,CAAC,CAAC;QAClC,IAAI,EAAE,IAAI,MAAM;YAAE,GAAG,CAAC,GAAG,CAAC,EAAE,EAAE,MAAM,CAAC,CAAC;IACxC,CAAC;IACD,OAAO,GAAG,CAAC;AACb,CAAC;AAED,SAAS,eAAe,CAAC,SAA6B;IACpD,MAAM,GAAG,GAAG,IAAI,GAAG,EAA0B,CAAC;IAC9C,IAAI,CAAC,SAAS;QAAE,OAAO,GAAG,CAAC;IAE3B,MAAM,OAAO,GAAG,sCAAsC,CAAC;IACvD,IAAI,EAA0B,CAAC;IAC/B,OAAO,CAAC,EAAE,GAAG,OAAO,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QAC/C,MAAM,KAAK,GAAG,EAAE,CAAC,CAAC,CAAC,CAAC;QACpB,MAAM,IAAI,GAAG,KAAK,CAAC,KAAK,CAAC,kBAAkB,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;QACxD,MAAM,IAAI,GAAG,IAAI,CAAC,IAAI,EAAE,QAAQ,CAAC,CAAC;QAClC,IAAI,IAAI,IAAI,IAAI,KAAK,WAAW,IAAI,IAAI,KAAK,WAAW,EAAE,CAAC;YACzD,+DAA+D;QACjE,CAAC;QACD,MAAM,OAAO,GACX,IAAI,CAAC,IAAI,EAAE,WAAW,CAAC;YACvB,+BAA+B,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC;QACnD,IAAI,CAAC,OAAO;YAAE,SAAS;QAEvB,MAAM,GAAG,GAAG,iCAAiC,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;QACrE,MAAM,KAAK,GAAmB,EAAE,CAAC;QACjC,IAAI,cAAc,CAAC,GAAG,EAAE,GAAG,CAAC,IAAI,cAAc,CAAC,GAAG,EAAE,KAAK,CAAC;YAAE,KAAK,CAAC,IAAI,GAAG,IAAI,CAAC;QAC9E,IAAI,cAAc,CAAC,GAAG,EAAE,GAAG,CAAC,IAAI,cAAc,CAAC,GAAG,EAAE,KAAK,CAAC;YAAE,KAAK,CAAC,MAAM,GAAG,IAAI,CAAC;QAChF,IAAI,cAAc,CAAC,GAAG,EAAE,QAAQ,CAAC,IAAI,cAAc,CAAC,GAAG,EAAE,SAAS,CAAC,EAAE,CAAC;YACpE,KAAK,CAAC,MAAM,GAAG,IAAI,CAAC;QACtB,CAAC;QAED,0EAA0E;QAC1E,IAAI,WAAW,CAAC,IAAI,CAAC,OAAO,CAAC;YAAE,KAAK,CAAC,IAAI,GAAG,IAAI,CAAC;QACjD,IAAI,8CAA8C,CAAC,IAAI,CAAC,OAAO,CAAC;YAAE,KAAK,CAAC,MAAM,GAAG,IAAI,CAAC;QACtF,IAAI,kBAAkB,CAAC,IAAI,CAAC,OAAO,CAAC;YAAE,KAAK,CAAC,IAAI,GAAG,IAAI,CAAC;QAExD,IAAI,KAAK,CAAC,IAAI,IAAI,KAAK,CAAC,MAAM,IAAI,KAAK,CAAC,MAAM;YAAE,GAAG,CAAC,GAAG,CAAC,OAAO,EAAE,KAAK,CAAC,CAAC;IAC1E,CAAC;IAED,uBAAuB;IACvB,MAAM,OAAO,GAAG,IAAI,GAAG,EAAkB,CAAC;IAC1C,MAAM,QAAQ,GAAG,sCAAsC,CAAC;IACxD,IAAI,GAA2B,CAAC;IAChC,OAAO,CAAC,GAAG,GAAG,QAAQ,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACjD,MAAM,KAAK,GAAG,GAAG,CAAC,CAAC,CAAC,CAAC;QACrB,MAAM,IAAI,GAAG,KAAK,CAAC,KAAK,CAAC,kBAAkB,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;QACxD,MAAM,OAAO,GAAG,IAAI,CAAC,IAAI,EAAE,WAAW,CAAC,CAAC;QACxC,MAAM,IAAI,GAAG,kCAAkC,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC;QACjE,IAAI,OAAO,IAAI,IAAI;YAAE,OAAO,CAAC,GAAG,CAAC,OAAO,EAAE,IAAI,CAAC,CAAC;IAClD,CAAC;IAED,MAAM,OAAO,GAAG,CAAC,EAAU,EAAE,OAAO,IAAI,GAAG,EAAU,EAAkB,EAAE;QACvE,IAAI,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC;YAAE,OAAO,GAAG,CAAC,GAAG,CAAC,EAAE,CAAC,IAAI,EAAE,CAAC;QAC3C,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;QACb,MAAM,GAAG,GAAG,GAAG,CAAC,GAAG,CAAC,EAAE,CAAC,IAAI,EAAE,CAAC;QAC9B,MAAM,QAAQ,GAAG,OAAO,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;QACjC,IAAI,CAAC,QAAQ;YAAE,OAAO,GAAG,CAAC;QAC1B,MAAM,MAAM,GAAG,OAAO,CAAC,QAAQ,EAAE,IAAI,CAAC,CAAC;QACvC,OAAO;YACL,IAAI,EAAE,GAAG,CAAC,IAAI,IAAI,MAAM,CAAC,IAAI;YAC7B,MAAM,EAAE,GAAG,CAAC,MAAM,IAAI,MAAM,CAAC,MAAM;YACnC,MAAM,EAAE,GAAG,CAAC,MAAM,IAAI,MAAM,CAAC,MAAM;SACpC,CAAC;IACJ,CAAC,CAAC;IAEF,KAAK,MAAM,EAAE,IAAI,IAAI,GAAG,CAAC,CAAC,GAAG,GAAG,CAAC,IAAI,EAAE,EAAE,GAAG,OAAO,CAAC,IAAI,EAAE,CAAC,CAAC,EAAE,CAAC;QAC7D,MAAM,QAAQ,GAAG,OAAO,CAAC,EAAE,CAAC,CAAC;QAC7B,IAAI,QAAQ,CAAC,IAAI,IAAI,QAAQ,CAAC,MAAM,IAAI,QAAQ,CAAC,MAAM;YAAE,GAAG,CAAC,GAAG,CAAC,EAAE,EAAE,QAAQ,CAAC,CAAC;IACjF,CAAC;IAED,OAAO,GAAG,CAAC;AACb,CAAC;AAED,SAAS,cAAc,CAAC,YAAgC;IACtD,MAAM,aAAa,GAAG,IAAI,GAAG,EAAkB,CAAC;IAChD,MAAM,cAAc,GAAG,IAAI,GAAG,EAA+B,CAAC;IAC9D,IAAI,CAAC,YAAY;QAAE,OAAO,EAAE,aAAa,EAAE,cAAc,EAAE,CAAC;IAE5D,MAAM,KAAK,GAAG,kDAAkD,CAAC;IACjE,IAAI,EAA0B,CAAC;IAC/B,OAAO,CAAC,EAAE,GAAG,KAAK,CAAC,IAAI,CAAC,YAAY,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QAChD,MAAM,KAAK,GAAG,EAAE,CAAC,CAAC,CAAC,CAAC;QACpB,MAAM,IAAI,GAAG,KAAK,CAAC,KAAK,CAAC,wBAAwB,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;QAC9D,MAAM,KAAK,GAAG,IAAI,CAAC,IAAI,EAAE,iBAAiB,CAAC,CAAC;QAC5C,IAAI,KAAK,IAAI,IAAI;YAAE,SAAS;QAC5B,MAAM,MAAM,GAAG,IAAI,GAAG,EAAkB,CAAC;QACzC,MAAM,KAAK,GAAG,kCAAkC,CAAC;QACjD,IAAI,EAA0B,CAAC;QAC/B,OAAO,CAAC,EAAE,GAAG,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;YACzC,MAAM,QAAQ,GAAG,EAAE,CAAC,CAAC,CAAC,CAAC;YACvB,MAAM,OAAO,GAAG,QAAQ,CAAC,KAAK,CAAC,gBAAgB,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;YAC5D,MAAM,IAAI,GAAG,MAAM,CAAC,IAAI,CAAC,OAAO,EAAE,QAAQ,CAAC,IAAI,GAAG,CAAC,CAAC;YACpD,MAAM,GAAG,GACP,iCAAiC,CAAC,IAAI,CAAC,QAAQ,CAAC,EAAE,CAAC,CAAC,CAAC;gBACrD,iCAAiC,CAAC,IAAI,CAAC,QAAQ,CAAC,EAAE,CAAC,CAAC,CAAC;gBACrD,QAAQ,CAAC;YACX,MAAM,CAAC,GAAG,CAAC,IAAI,EAAE,GAAG,CAAC,CAAC;QACxB,CAAC;QACD,cAAc,CAAC,GAAG,CAAC,KAAK,EAAE,MAAM,CAAC,CAAC;IACpC,CAAC;IAED,MAAM,KAAK,GAAG,kCAAkC,CAAC;IACjD,IAAI,EAA0B,CAAC;IAC/B,OAAO,CAAC,EAAE,GAAG,KAAK,CAAC,IAAI,CAAC,YAAY,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QAChD,MAAM,KAAK,GAAG,EAAE,CAAC,CAAC,CAAC,CAAC;QACpB,MAAM,IAAI,GAAG,KAAK,CAAC,KAAK,CAAC,gBAAgB,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;QACtD,MAAM,KAAK,GAAG,IAAI,CAAC,IAAI,EAAE,SAAS,CAAC,CAAC;QACpC,MAAM,KAAK,GACT,wCAAwC,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC,CAAC;YACzD,wCAAwC,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC;QAC5D,IAAI,KAAK,IAAI,IAAI,IAAI,KAAK,IAAI,IAAI;YAAE,aAAa,CAAC,GAAG,CAAC,KAAK,EAAE,KAAK,CAAC,CAAC;IACtE,CAAC;IAED,OAAO,EAAE,aAAa,EAAE,cAAc,EAAE,CAAC;AAC3C,CAAC;AAED,SAAS,aAAa,CACpB,OAAe,EACf,SAAyB;IAEzB,MAAM,KAAK,GACT,gCAAgC,CAAC,IAAI,CAAC,OAAO,CAAC,EAAE,CAAC,CAAC,CAAC;QACnD,gCAAgC,CAAC,IAAI,CAAC,OAAO,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC;IACtD,IAAI,KAAK,IAAI,IAAI;QAAE,OAAO,KAAK,CAAC;IAChC,MAAM,KAAK,GAAG,SAAS,CAAC,aAAa,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC;IACjD,IAAI,KAAK,IAAI,IAAI;QAAE,OAAO,KAAK,CAAC;IAChC,MAAM,IAAI,GAAG,SAAS,CAAC,OAAO,CAAC,CAAC;IAChC,MAAM,GAAG,GAAG,SAAS,CAAC,cAAc,CAAC,GAAG,CAAC,KAAK,CAAC,EAAE,GAAG,CAAC,IAAI,CAAC,IAAI,QAAQ,CAAC;IACvE,OAAO,GAAG,KAAK,QAAQ,IAAI,GAAG,KAAK,MAAM,CAAC;AAC5C,CAAC;AAED,SAAS,eAAe,CAAC,KAAa;IACpC,MAAM,KAAK,GAAa,EAAE,CAAC;IAC3B,6EAA6E;IAC7E,MAAM,OAAO,GAAG,8EAA8E,CAAC;IAC/F,IAAI,EAA0B,CAAC;IAC/B,OAAO,CAAC,EAAE,GAAG,OAAO,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QAC3C,IAAI,EAAE,CAAC,CAAC,CAAC,CAAC,UAAU,CAAC,QAAQ,CAAC;YAAE,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;;YAC5C,KAAK,CAAC,IAAI,CAAC,iBAAiB,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC;IAClD,CAAC;IACD,OAAO,KAAK,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;AACxB,CAAC;AAED,SAAS,cAAc,CAAC,GAAW,EAAE,GAAW;IAC9C,8EAA8E;IAC9E,MAAM,EAAE,GAAG,IAAI,MAAM,CAAC,MAAM,GAAG,aAAa,EAAE,GAAG,CAAC,CAAC;IACnD,MAAM,CAAC,GAAG,EAAE,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;IACvB,IAAI,CAAC,CAAC;QAAE,OAAO,KAAK,CAAC;IACrB,MAAM,KAAK,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;IACnB,MAAM,GAAG,GACP,kBAAkB,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,kBAAkB,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC;IAC7E,IAAI,GAAG,IAAI,IAAI,EAAE,CAAC;QAChB,MAAM,CAAC,GAAG,GAAG,CAAC,WAAW,EAAE,CAAC;QAC5B,OAAO,CAAC,KAAK,GAAG,IAAI,CAAC,KAAK,OAAO,IAAI,CAAC,KAAK,KAAK,CAAC;IACnD,CAAC;IACD,OAAO,IAAI,CAAC;AACd,CAAC;AAED,SAAS,YAAY,CACnB,KAAa,EACb,UAAuC;IAEvC,MAAM,GAAG,GAAG,iCAAiC,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;IACrE,MAAM,OAAO,GACX,iCAAiC,CAAC,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,CAAC;QAChD,iCAAiC,CAAC,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC;IACnD,MAAM,SAAS,GAAG,OAAO,CAAC,CAAC,CAAC,UAAU,CAAC,GAAG,CAAC,OAAO,CAAC,IAAI,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;IAE/D,MAAM,IAAI,GACR,SAAS,CAAC,IAAI,IAAI,cAAc,CAAC,GAAG,EAAE,GAAG,CAAC,IAAI,cAAc,CAAC,GAAG,EAAE,KAAK,CAAC,CAAC;IAC3E,MAAM,MAAM,GACV,SAAS,CAAC,MAAM,IAAI,cAAc,CAAC,GAAG,EAAE,GAAG,CAAC,IAAI,cAAc,CAAC,GAAG,EAAE,KAAK,CAAC,CAAC;IAC7E,MAAM,MAAM,GACV,SAAS,CAAC,MAAM;QAChB,cAAc,CAAC,GAAG,EAAE,QAAQ,CAAC;QAC7B,cAAc,CAAC,GAAG,EAAE,SAAS,CAAC,CAAC;IAEjC,OAAO;QACL,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QAC/B,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;QACnC,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;KACpC,CAAC;AACJ,CAAC;AAED,SAAS,eAAe,CACtB,MAAc,EACd,UAAuC,EACvC,IAAa;IAEb,6CAA6C;IAC7C,MAAM,EAAE,GACN,2CAA2C,CAAC,IAAI,CAAC,MAAM,CAAC,EAAE,CAAC,CAAC,CAAC;QAC7D,2CAA2C,CAAC,IAAI,CAAC,MAAM,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC;IAChE,IAAI,EAAE,IAAI,IAAI,IAAI,EAAE,KAAK,GAAG,IAAI,EAAE,KAAK,IAAI,EAAE,CAAC;QAC5C,OAAO,EAAE,IAAI,EAAE,OAAO,EAAE,GAAG,EAAE,CAAC;IAChC,CAAC;IACD,MAAM,EAAE,GACN,0CAA0C,CAAC,IAAI,CAAC,MAAM,CAAC,EAAE,CAAC,CAAC,CAAC;QAC5D,0CAA0C,CAAC,IAAI,CAAC,MAAM,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC;IAC/D,IAAI,EAAE,IAAI,IAAI,IAAI,EAAE,KAAK,GAAG,IAAI,EAAE,KAAK,IAAI,EAAE,CAAC;QAC5C,OAAO,EAAE,IAAI,EAAE,OAAO,EAAE,GAAG,EAAE,CAAC;IAChC,CAAC;IAED,MAAM,IAAI,GAAG,eAAe,CAAC,MAAM,CAAC,CAAC;IACrC,IAAI,CAAC,IAAI;QAAE,OAAO,IAAI,CAAC;IACvB,MAAM,KAAK,GAAG,YAAY,CAAC,MAAM,EAAE,UAAU,CAAC,CAAC;IAC/C,OAAO;QACL,IAAI;QACJ,GAAG,KAAK;QACR,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;KAC1B,CAAC;AACJ,CAAC;AAED;;GAEG;AACH,SAAS,kBAAkB,CACzB,OAAe,EACf,UAAuC,EACvC,IAAyB;IAEzB,MAAM,IAAI,GAAgB,EAAE,CAAC;IAC7B,2CAA2C;IAC3C,MAAM,UAAU,GAAG,OAAO,CAAC,OAAO,CAAC,iCAAiC,EAAE,EAAE,CAAC,CAAC;IAE1E,MAAM,OAAO,GAAG,0EAA0E,CAAC;IAC3F,IAAI,CAAyB,CAAC;IAC9B,OAAO,CAAC,CAAC,GAAG,OAAO,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QAC/C,MAAM,KAAK,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;QACnB,IAAI,KAAK,CAAC,UAAU,CAAC,cAAc,CAAC,EAAE,CAAC;YACrC,MAAM,IAAI,GAAG,KAAK,CAAC,KAAK,CAAC,sBAAsB,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;YAC5D,MAAM,GAAG,GAAG,IAAI,CAAC,IAAI,EAAE,MAAM,CAAC,CAAC;YAC/B,MAAM,MAAM,GAAG,IAAI,CAAC,IAAI,EAAE,UAAU,CAAC,CAAC;YACtC,IAAI,IAAwB,CAAC;YAC7B,IAAI,GAAG,IAAI,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC;gBAAE,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;iBAC1C,IAAI,MAAM;gBAAE,IAAI,GAAG,IAAI,MAAM,EAAE,CAAC;YAErC,MAAM,UAAU,GAAG,8BAA8B,CAAC;YAClD,IAAI,EAA0B,CAAC;YAC/B,OAAO,CAAC,EAAE,GAAG,UAAU,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;gBAC9C,MAAM,GAAG,GAAG,eAAe,CAAC,EAAE,CAAC,CAAC,CAAC,EAAE,UAAU,EAAE,IAAI,CAAC,CAAC;gBACrD,IAAI,GAAG;oBAAE,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;YAC1B,CAAC;QACH,CAAC;aAAM,CAAC;YACN,MAAM,GAAG,GAAG,eAAe,CAAC,KAAK,EAAE,UAAU,CAAC,CAAC;YAC/C,IAAI,GAAG;gBAAE,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QAC1B,CAAC;IACH,CAAC;IAED,IAAI,IAAI,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QACtB,MAAM,IAAI,GAAG,eAAe,CAAC,UAAU,CAAC,CAAC;QACzC,oCAAoC;QACpC,MAAM,OAAO,GAAG,IAAI,CAAC,OAAO,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC;QACrE,IAAI,OAAO;YAAE,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,OAAO,EAAE,CAAC,CAAC;IAC5C,CAAC;IACD,OAAO,IAAI,CAAC;AACd,CAAC;AAED,SAAS,gBAAgB,CAAC,OAAe;IACvC,OAAO,CACL,iCAAiC,CAAC,IAAI,CAAC,OAAO,CAAC,EAAE,CAAC,CAAC,CAAC;QACpD,iCAAiC,CAAC,IAAI,CAAC,OAAO,CAAC,EAAE,CAAC,CAAC,CAAC,CACrD,CAAC;AACJ,CAAC;AAED,SAAS,YAAY,CAAC,OAAe;IACnC,MAAM,KAAK,GAAG,gBAAgB,CAAC,OAAO,CAAC,CAAC;IACxC,IAAI,KAAK,EAAE,CAAC;QACV,IAAI,UAAU,CAAC,IAAI,CAAC,KAAK,CAAC;YAAE,OAAO,CAAC,CAAC;QACrC,MAAM,EAAE,GAAG,sBAAsB,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,sBAAsB,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QACpF,IAAI,EAAE;YAAE,OAAO,MAAM,CAAC,EAAE,CAAC,CAAC,CAAC,CAA0B,CAAC;IACxD,CAAC;IACD,MAAM,OAAO,GAAG,mCAAmC,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;IAClE,IAAI,OAAO,EAAE,CAAC;QACZ,MAAM,GAAG,GAAG,MAAM,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC;QACnC,IAAI,GAAG,IAAI,CAAC,IAAI,GAAG,IAAI,CAAC;YAAE,OAAO,GAA4B,CAAC;IAChE,CAAC;IACD,OAAO,IAAI,CAAC;AACd,CAAC;AAED,SAAS,QAAQ,CAAC,OAAe;IAC/B,MAAM,KAAK,GAAG,gBAAgB,CAAC,OAAO,CAAC,CAAC;IACxC,IAAI,CAAC,KAAK;QAAE,OAAO,IAAI,CAAC;IACxB,IAAI,eAAe,CAAC,IAAI,CAAC,KAAK,CAAC;QAAE,OAAO,CAAC,CAAC;IAC1C,MAAM,CAAC,GAAG,eAAe,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;IACtC,IAAI,CAAC;QAAE,OAAO,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC;IAC/B,OAAO,IAAI,CAAC;AACd,CAAC;AAED,SAAS,UAAU,CAAC,OAAe;IACjC,OAAO,6BAA6B,CAAC,IAAI,CAAC,OAAO,CAAC,IAAI,oBAAoB,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;AAC3F,CAAC;AAED,SAAS,SAAS,CAAC,OAAe;IAChC,MAAM,CAAC,GAAG,6BAA6B,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;IACtD,OAAO,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;AAC9B,CAAC;AAED,SAAS,SAAS,CAAC,OAAe;IAChC,OAAO,aAAa,CAAC,IAAI,CAAC,OAAO,CAAC,IAAI,aAAa,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;AACpE,CAAC;AAED,6FAA6F;AAC7F,SAAS,gBAAgB,CAAC,IAAgB;IACxC,OAAO,IAAI;SACR,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;SACnF,MAAM,CAAC,OAAO,CAAC;SACf,IAAI,CAAC,KAAK,CAAC,CAAC;AACjB,CAAC;AAED,SAAS,cAAc,CAAC,MAAc;IACpC,MAAM,IAAI,GAAe,EAAE,CAAC;IAC5B,wDAAwD;IACxD,MAAM,KAAK,GAAG,MAAM,CAAC,OAAO,CAAC,iBAAiB,EAAE,EAAE,CAAC,CAAC,OAAO,CAAC,YAAY,EAAE,EAAE,CAAC,CAAC;IAC9E,IAAI,CAAC,GAAG,CAAC,CAAC;IACV,OAAO,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC;QACxB,MAAM,IAAI,GAAG,KAAK,CAAC,OAAO,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC;QACvC,IAAI,IAAI,GAAG,CAAC;YAAE,MAAM;QACpB,MAAM,KAAK,GAAG,KAAK,CAAC,MAAM,CAAC,IAAI,GAAG,CAAC,CAAC,CAAC;QACrC,IAAI,KAAK,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC;YACnC,CAAC,GAAG,IAAI,GAAG,CAAC,CAAC;YACb,SAAS;QACX,CAAC;QACD,MAAM,KAAK,GAAG,eAAe,CAAC,KAAK,EAAE,IAAI,EAAE,IAAI,CAAC,CAAC;QACjD,IAAI,KAAK,GAAG,CAAC,EAAE,CAAC;YACd,CAAC,GAAG,IAAI,GAAG,CAAC,CAAC;YACb,SAAS;QACX,CAAC;QACD,MAAM,KAAK,GAAG,KAAK,CAAC,KAAK,CAAC,IAAI,EAAE,KAAK,CAAC,CAAC;QACvC,MAAM,KAAK,GAAa,EAAE,CAAC;QAC3B,IAAI,CAAC,GAAG,CAAC,CAAC;QACV,MAAM,OAAO,GAAG,KAAK,CAAC;QACtB,OAAO,CAAC,GAAG,OAAO,CAAC,MAAM,EAAE,CAAC;YAC1B,MAAM,IAAI,GAAG,OAAO,CAAC,OAAO,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC;YACzC,IAAI,IAAI,GAAG,CAAC;gBAAE,MAAM;YACpB,MAAM,OAAO,GAAG,OAAO,CAAC,MAAM,CAAC,IAAI,GAAG,CAAC,CAAC,CAAC;YACzC,IAAI,OAAO,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,OAAO,CAAC,EAAE,CAAC;gBACvC,CAAC,GAAG,IAAI,GAAG,CAAC,CAAC;gBACb,SAAS;YACX,CAAC;YACD,MAAM,KAAK,GAAG,eAAe,CAAC,OAAO,EAAE,IAAI,EAAE,IAAI,CAAC,CAAC;YACnD,IAAI,KAAK,GAAG,CAAC,EAAE,CAAC;gBACd,CAAC,GAAG,IAAI,GAAG,CAAC,CAAC;gBACb,SAAS;YACX,CAAC;YACD,KAAK,CAAC,IAAI,CAAC,aAAa,CAAC,OAAO,CAAC,KAAK,CAAC,IAAI,EAAE,KAAK,CAAC,CAAC,CAAC,CAAC;YACtD,CAAC,GAAG,KAAK,CAAC;QACZ,CAAC;QACD,IAAI,KAAK,CAAC,MAAM;YAAE,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QACnC,CAAC,GAAG,KAAK,CAAC;IACZ,CAAC;IACD,OAAO,IAAI,CAAC;AACd,CAAC;AAED,SAAS,aAAa,CAAC,KAAa;IAClC,MAAM,KAAK,GAAa,EAAE,CAAC;IAC3B,cAAc;IACd,IAAI,IAAI,GAAG,KAAK,CAAC,OAAO,CAAC,mCAAmC,EAAE,EAAE,CAAC,CAAC;IAClE,IAAI,CAAC,GAAG,CAAC,CAAC;IACV,OAAO,CAAC,GAAG,IAAI,CAAC,MAAM,EAAE,CAAC;QACvB,MAAM,KAAK,GAAG,IAAI,CAAC,OAAO,CAAC,QAAQ,EAAE,CAAC,CAAC,CAAC;QACxC,MAAM,GAAG,GAAG,IAAI,CAAC,OAAO,CAAC,MAAM,EAAE,CAAC,CAAC,CAAC;QACpC,IAAI,KAAK,GAAG,CAAC,IAAI,GAAG,GAAG,CAAC;YAAE,MAAM;QAChC,IAAI,KAAK,IAAI,CAAC,IAAI,CAAC,GAAG,GAAG,CAAC,IAAI,KAAK,GAAG,GAAG,CAAC,EAAE,CAAC;YAC3C,MAAM,KAAK,GAAG,IAAI,CAAC,MAAM,CAAC,KAAK,GAAG,CAAC,CAAC,CAAC;YACrC,IAAI,KAAK,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC;gBACnC,CAAC,GAAG,KAAK,GAAG,CAAC,CAAC;gBACd,SAAS;YACX,CAAC;YACD,MAAM,GAAG,GAAG,eAAe,CAAC,IAAI,EAAE,KAAK,EAAE,KAAK,CAAC,CAAC;YAChD,IAAI,GAAG,GAAG,CAAC,EAAE,CAAC;gBACZ,CAAC,GAAG,KAAK,GAAG,CAAC,CAAC;gBACd,SAAS;YACX,CAAC;YACD,MAAM,UAAU,GAAG,cAAc,CAAC,IAAI,CAAC,KAAK,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,CAAC;YAC1D,MAAM,OAAO,GAAG,gBAAgB,CAAC,UAAU,CAAC,CAAC;YAC7C,IAAI,OAAO;gBAAE,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;YACjC,CAAC,GAAG,GAAG,CAAC;QACV,CAAC;aAAM,CAAC;YACN,MAAM,KAAK,GAAG,IAAI,CAAC,MAAM,CAAC,GAAG,GAAG,CAAC,CAAC,CAAC;YACnC,IAAI,KAAK,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,KAAK,CAAC,EAAE,CAAC;gBACnC,CAAC,GAAG,GAAG,GAAG,CAAC,CAAC;gBACZ,SAAS;YACX,CAAC;YACD,MAAM,GAAG,GAAG,eAAe,CAAC,IAAI,EAAE,GAAG,EAAE,GAAG,CAAC,CAAC;YAC5C,IAAI,GAAG,GAAG,CAAC,EAAE,CAAC;gBACZ,CAAC,GAAG,GAAG,GAAG,CAAC,CAAC;gBACZ,SAAS;YACX,CAAC;YACD,MAAM,IAAI,GAAG,eAAe,CAAC,IAAI,CAAC,KAAK,CAAC,GAAG,EAAE,GAAG,CAAC,CAAC,CAAC,OAAO,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC;YAC9E,IAAI,IAAI;gBAAE,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;YAC3B,CAAC,GAAG,GAAG,CAAC;QACV,CAAC;IACH,CAAC;IACD,OAAO,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC;AAChC,CAAC;AAED,SAAS,UAAU,CAAC,MAAc;IAChC,OAAO,EAAE,IAAI,EAAE,OAAO,EAAE,IAAI,EAAE,cAAc,CAAC,MAAM,CAAC,EAAE,CAAC;AACzD,CAAC;AAQD,SAAS,WAAW,CAAC,OAAe,EAAE,GAAa;IACjD,MAAM,IAAI,GAAG,kBAAkB,CAAC,OAAO,EAAE,GAAG,CAAC,UAAU,EAAE,GAAG,CAAC,IAAI,CAAC,CAAC;IACnE,gDAAgD;IAChD,MAAM,MAAM,GAAG,QAAQ,CAAC,OAAO,CAAC,CAAC;IACjC,IAAI,MAAM,IAAI,IAAI,EAAE,CAAC;QACnB,MAAM,KAAK,GAAG,IAAI;aACf,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC;aAClB,IAAI,CAAC,EAAE,CAAC;aACR,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;aACd,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC;aACpB,IAAI,EAAE,CAAC;QACV,IAAI,CAAC,KAAK;YAAE,OAAO,IAAI,CAAC;QACxB,IAAI,eAAe,CAAC,IAAI,CAAC,gBAAgB,CAAC,OAAO,CAAC,IAAI,EAAE,CAAC,EAAE,CAAC;YAC1D,OAAO,EAAE,IAAI,EAAE,SAAS,EAAE,KAAK,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,KAAK,EAAE,CAAC,EAAE,CAAC;QAChE,CAAC;QACD,OAAO;YACL,IAAI,EAAE,WAAW;YACjB,KAAK,EAAE,MAAM;YACb,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,KAAK,EAAE,CAAC;SACxB,CAAC;IACJ,CAAC;IAED,MAAM,KAAK,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC;IACtD,IAAI,CAAC,KAAK,IAAI,CAAC,UAAU,CAAC,OAAO,CAAC;QAAE,OAAO,IAAI,CAAC;IAEhD,MAAM,KAAK,GAAG,YAAY,CAAC,OAAO,CAAC,CAAC;IACpC,IAAI,KAAK,EAAE,CAAC;QACV,OAAO,EAAE,IAAI,EAAE,SAAS,EAAE,KAAK,EAAE,IAAI,EAAE,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,KAAK,EAAE,CAAC,EAAE,CAAC;IAClF,CAAC;IACD,IAAI,UAAU,CAAC,OAAO,CAAC,EAAE,CAAC;QACxB,MAAM,OAAO,GAAG,aAAa,CAAC,OAAO,EAAE,GAAG,CAAC,SAAS,CAAC,CAAC;QACtD,OAAO;YACL,IAAI,EAAE,WAAW;YACjB,KAAK,EAAE,SAAS,CAAC,OAAO,CAAC;YACzB,GAAG,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,OAAO,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;YACrC,IAAI,EAAE,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,KAAK,IAAI,GAAG,EAAE,CAAC;SACpD,CAAC;IACJ,CAAC;IACD,IAAI,CAAC,KAAK;QAAE,OAAO,IAAI,CAAC;IACxB,OAAO,EAAE,IAAI,EAAE,WAAW,EAAE,IAAI,EAAE,CAAC;AACrC,CAAC;AAED,SAAS,aAAa,CACpB,GAAuB,EACvB,IAA4B;IAE5B,IAAI,CAAC,GAAG;QAAE,OAAO,EAAE,CAAC;IACpB,MAAM,GAAG,GAAG,IAAI,KAAK,UAAU,CAAC,CAAC,CAAC,UAAU,CAAC,CAAC,CAAC,SAAS,CAAC;IACzD,qEAAqE;IACrE,MAAM,EAAE,GAAG,IAAI,MAAM,CACnB,MAAM,GAAG,yCAAyC,GAAG,GAAG,EACxD,GAAG,CACJ,CAAC;IACF,MAAM,GAAG,GAAwC,EAAE,CAAC;IACpD,IAAI,CAAyB,CAAC;IAC9B,OAAO,CAAC,CAAC,GAAG,EAAE,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACnC,MAAM,IAAI,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC,CAAC;QAClD,MAAM,IAAI,GAAG,IAAI,CAAC,IAAI,EAAE,QAAQ,CAAC,CAAC;QAClC,IAAI,IAAI,KAAK,WAAW,IAAI,IAAI,KAAK,uBAAuB;YAAE,SAAS;QACvE,MAAM,EAAE,GAAG,IAAI,CAAC,IAAI,EAAE,MAAM,CAAC,CAAC;QAC9B,IAAI,EAAE,IAAI,IAAI,IAAI,EAAE,KAAK,IAAI,IAAI,EAAE,KAAK,GAAG;YAAE,SAAS;QACtD,MAAM,IAAI,GAAG,eAAe,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC;QAC/D,IAAI,IAAI;YAAE,GAAG,CAAC,IAAI,CAAC,EAAE,EAAE,EAAE,IAAI,EAAE,CAAC,CAAC;IACnC,CAAC;IACD,OAAO,GAAG,CAAC;AACb,CAAC;AAED,uDAAuD;AACvD,MAAM,UAAU,0BAA0B,CAAC,KAAuB;IAChE,MAAM,GAAG,GAAa;QACpB,UAAU,EAAE,eAAe,CAAC,KAAK,CAAC,SAAS,CAAC;QAC5C,SAAS,EAAE,cAAc,CAAC,KAAK,CAAC,YAAY,CAAC;QAC7C,IAAI,EAAE,SAAS,CAAC,KAAK,CAAC,OAAO,CAAC;KAC/B,CAAC;IAEF,MAAM,IAAI,GAAG,gBAAgB,CAAC,KAAK,CAAC,WAAW,CAAC,CAAC;IACjD,MAAM,MAAM,GAAmB,EAAE,CAAC;IAClC,IAAI,UAAU,GAAG,EAAE,CAAC;IAEpB,KAAK,MAAM,IAAI,IAAI,aAAa,CAAC,IAAI,CAAC,EAAE,CAAC;QACvC,IAAI,IAAI,CAAC,IAAI,KAAK,KAAK,EAAE,CAAC;YACxB,UAAU,GAAG,EAAE,CAAC;YAChB,MAAM,KAAK,GAAG,UAAU,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;YACnC,IAAI,KAAK,CAAC,IAAI,KAAK,OAAO,IAAI,KAAK,CAAC,IAAI,CAAC,MAAM;gBAAE,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;YACpE,SAAS;QACX,CAAC;QAED,IAAI,SAAS,CAAC,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC;YACxB,MAAM,MAAM,GAAG,eAAe,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC;YAChD,UAAU,IAAI,MAAM,CAAC;YACrB,SAAS;QACX,CAAC;QAED,MAAM,KAAK,GAAG,WAAW,CAAC,IAAI,CAAC,GAAG,EAAE,GAAG,CAAC,CAAC;QACzC,IAAI,CAAC,KAAK;YAAE,SAAS;QAErB,IAAI,UAAU,IAAI,CAAC,KAAK,CAAC,IAAI,KAAK,WAAW,IAAI,KAAK,CAAC,IAAI,KAAK,WAAW,CAAC,EAAE,CAAC;YAC7E,MAAM,IAAI,GAAG,CAAC,GAAG,KAAK,CAAC,IAAI,CAAC,CAAC;YAC7B,IAAI,IAAI,CAAC,MAAM,EAAE,CAAC;gBAChB,IAAI,CAAC,CAAC,CAAC,GAAG,EAAE,GAAG,IAAI,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,UAAU,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;YAC5D,CAAC;iBAAM,CAAC;gBACN,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,UAAU,EAAE,CAAC,CAAC;YAClC,CAAC;YACD,UAAU,GAAG,EAAE,CAAC;YAChB,MAAM,CAAC,IAAI,CAAC,EAAE,GAAG,KAAK,EAAE,IAAI,EAAE,CAAC,CAAC;QAClC,CAAC;aAAM,CAAC;YACN,UAAU,GAAG,EAAE,CAAC;YAChB,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QACrB,CAAC;IACH,CAAC;IAED,MAAM,SAAS,GAAG,aAAa,CAAC,KAAK,CAAC,YAAY,EAAE,UAAU,CAAC,CAAC;IAChE,MAAM,QAAQ,GAAG,aAAa,CAAC,KAAK,CAAC,WAAW,EAAE,SAAS,CAAC,CAAC;IAC7D,IAAI,SAAS,CAAC,MAAM,IAAI,QAAQ,CAAC,MAAM,EAAE,CAAC;QACxC,KAAK,MAAM,EAAE,IAAI,SAAS,EAAE,CAAC;YAC3B,MAAM,CAAC,IAAI,CAAC;gBACV,IAAI,EAAE,WAAW;gBACjB,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,OAAO,EAAE,CAAC,EAAE,MAAM,EAAE,CAAC,IAAI,EAAE,EAAE,CAAC;aAC9C,CAAC,CAAC;QACL,CAAC;QACD,KAAK,MAAM,EAAE,IAAI,QAAQ,EAAE,CAAC;YAC1B,MAAM,CAAC,IAAI,CAAC;gBACV,IAAI,EAAE,WAAW;gBACjB,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,OAAO,EAAE,CAAC,EAAE,MAAM,EAAE,CAAC,IAAI,EAAE,EAAE,CAAC;aAC9C,CAAC,CAAC;QACL,CAAC;IACH,CAAC;IAED,OAAO,MAAM,CAAC;AAChB,CAAC;AAED,sEAAsE;AACtE,MAAM,UAAU,sBAAsB,CAAC,GAAW;IAChD,OAAO,0BAA0B,CAAC,EAAE,WAAW,EAAE,GAAG,EAAE,CAAC,CAAC;AAC1D,CAAC;AAED,MAAM,UAAU,yBAAyB,CAAC,KAAuB;IAC/D,MAAM,MAAM,GAAG,0BAA0B,CAAC,KAAK,CAAC,CAAC;IACjD,OAAO;QACL,MAAM,EAAE,MAAM;QACd,MAAM;QACN,UAAU,EAAE,yBAAyB,CAAC,MAAM,CAAC;KAC9C,CAAC;AACJ,CAAC;AAED,MAAM,UAAU,qBAAqB,CAAC,GAAW;IAC/C,OAAO,yBAAyB,CAAC,EAAE,WAAW,EAAE,GAAG,EAAE,CAAC,CAAC;AACzD,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,qBAAqB,CAAC,GAAU;IACpD,MAAM,OAAO,GAAG,GAAG,CAAC,IAAI,CAAC,mBAAmB,CAAC,CAAC;IAC9C,IAAI,CAAC,OAAO,EAAE,CAAC;QACb,MAAM,IAAI,aAAa,CAAC,cAAc,EAAE,yCAAyC,CAAC,CAAC;IACrF,CAAC;IACD,MAAM,WAAW,GAAG,MAAM,OAAO,CAAC,KAAK,CAAC,QAAQ,CAAC,CAAC;IAClD,MAAM,IAAI,GAAG,KAAK,EAAE,IAAY,EAAE,EAAE;QAClC,MAAM,CAAC,GAAG,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QACzB,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC;IAC3C,CAAC,CAAC;IACF,OAAO,yBAAyB,CAAC;QAC/B,WAAW;QACX,SAAS,EAAE,MAAM,IAAI,CAAC,iBAAiB,CAAC;QACxC,YAAY,EAAE,MAAM,IAAI,CAAC,oBAAoB,CAAC;QAC9C,OAAO,EAAE,MAAM,IAAI,CAAC,8BAA8B,CAAC;QACnD,YAAY,EAAE,MAAM,IAAI,CAAC,oBAAoB,CAAC;QAC9C,WAAW,EAAE,MAAM,IAAI,CAAC,mBAAmB,CAAC;KAC7C,CAAC,CAAC;AACL,CAAC"}
|
|
@@ -0,0 +1,14 @@
|
|
|
1
|
+
import type { ContentBlock } from '../types.js';
|
|
2
|
+
/**
|
|
3
|
+
* Convert HTML snippet into ContentBlocks (structural convert path).
|
|
4
|
+
*/
|
|
5
|
+
export declare function htmlToBlocks(html: string): ContentBlock[];
|
|
6
|
+
export declare function htmlToStructured(html: string): {
|
|
7
|
+
format: 'html';
|
|
8
|
+
paragraphs: string[];
|
|
9
|
+
blocks: ContentBlock[];
|
|
10
|
+
text: string;
|
|
11
|
+
};
|
|
12
|
+
/** Lightweight MD → blocks (headings, lists, GFM tables, paragraphs). */
|
|
13
|
+
export declare function markdownToBlocks(md: string): ContentBlock[];
|
|
14
|
+
//# sourceMappingURL=html-blocks.d.ts.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"html-blocks.d.ts","sourceRoot":"","sources":["../../src/parse/html-blocks.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,YAAY,EAAa,MAAM,aAAa,CAAC;AAiE3D;;GAEG;AACH,wBAAgB,YAAY,CAAC,IAAI,EAAE,MAAM,GAAG,YAAY,EAAE,CA6DzD;AAMD,wBAAgB,gBAAgB,CAAC,IAAI,EAAE,MAAM,GAAG;IAC9C,MAAM,EAAE,MAAM,CAAC;IACf,UAAU,EAAE,MAAM,EAAE,CAAC;IACrB,MAAM,EAAE,YAAY,EAAE,CAAC;IACvB,IAAI,EAAE,MAAM,CAAC;CACd,CASA;AAED,yEAAyE;AACzE,wBAAgB,gBAAgB,CAAC,EAAE,EAAE,MAAM,GAAG,YAAY,EAAE,CAqF3D"}
|
|
@@ -0,0 +1,244 @@
|
|
|
1
|
+
import { flattenBlocksToParagraphs } from '../convert/blocks.js';
|
|
2
|
+
function decodeEntities(s) {
|
|
3
|
+
return s
|
|
4
|
+
.replace(/ /gi, ' ')
|
|
5
|
+
.replace(/</gi, '<')
|
|
6
|
+
.replace(/>/gi, '>')
|
|
7
|
+
.replace(/"/gi, '"')
|
|
8
|
+
.replace(/'/gi, "'")
|
|
9
|
+
.replace(/&/gi, '&');
|
|
10
|
+
}
|
|
11
|
+
function parseInline(html) {
|
|
12
|
+
const runs = [];
|
|
13
|
+
const re = /<(strong|b)(\s[^>]*)?>((?:[\s\S]*?))<\/\1>|<(em|i)(\s[^>]*)?>((?:[\s\S]*?))<\/\4>|([^<]+)/gi;
|
|
14
|
+
let m;
|
|
15
|
+
while ((m = re.exec(html)) !== null) {
|
|
16
|
+
if (m[1]) {
|
|
17
|
+
const inner = m[3];
|
|
18
|
+
// Nested em/i inside strong/b
|
|
19
|
+
const nested = /<(em|i)(\s[^>]*)?>([\s\S]*?)<\/\1>/i.exec(inner);
|
|
20
|
+
if (nested) {
|
|
21
|
+
const before = decodeEntities(inner.slice(0, nested.index).replace(/<[^>]+>/g, ''));
|
|
22
|
+
if (before)
|
|
23
|
+
runs.push({ text: before, bold: true });
|
|
24
|
+
const mid = decodeEntities(nested[3].replace(/<[^>]+>/g, ''));
|
|
25
|
+
if (mid)
|
|
26
|
+
runs.push({ text: mid, bold: true, italic: true });
|
|
27
|
+
const after = decodeEntities(inner.slice(nested.index + nested[0].length).replace(/<[^>]+>/g, ''));
|
|
28
|
+
if (after)
|
|
29
|
+
runs.push({ text: after, bold: true });
|
|
30
|
+
}
|
|
31
|
+
else {
|
|
32
|
+
const text = decodeEntities(inner.replace(/<[^>]+>/g, ''));
|
|
33
|
+
if (text)
|
|
34
|
+
runs.push({ text, bold: true });
|
|
35
|
+
}
|
|
36
|
+
}
|
|
37
|
+
else if (m[4]) {
|
|
38
|
+
const text = decodeEntities(m[6].replace(/<[^>]+>/g, ''));
|
|
39
|
+
if (text)
|
|
40
|
+
runs.push({ text, italic: true });
|
|
41
|
+
}
|
|
42
|
+
else if (m[7]) {
|
|
43
|
+
const t = decodeEntities(m[7]);
|
|
44
|
+
if (t)
|
|
45
|
+
runs.push({ text: t });
|
|
46
|
+
}
|
|
47
|
+
}
|
|
48
|
+
if (!runs.length) {
|
|
49
|
+
const plain = decodeEntities(html.replace(/<[^>]+>/g, ''));
|
|
50
|
+
if (plain)
|
|
51
|
+
runs.push({ text: plain });
|
|
52
|
+
}
|
|
53
|
+
return runs;
|
|
54
|
+
}
|
|
55
|
+
function parseTableHtml(tableHtml) {
|
|
56
|
+
const rows = [];
|
|
57
|
+
const trRe = /<tr\b[^>]*>([\s\S]*?)<\/tr>/gi;
|
|
58
|
+
let tr;
|
|
59
|
+
while ((tr = trRe.exec(tableHtml)) !== null) {
|
|
60
|
+
const cells = [];
|
|
61
|
+
const cellRe = /<(td|th)\b[^>]*>([\s\S]*?)<\/\1>/gi;
|
|
62
|
+
let c;
|
|
63
|
+
while ((c = cellRe.exec(tr[1])) !== null) {
|
|
64
|
+
cells.push(decodeEntities(c[2].replace(/<[^>]+>/g, '')).trim());
|
|
65
|
+
}
|
|
66
|
+
if (cells.length)
|
|
67
|
+
rows.push(cells);
|
|
68
|
+
}
|
|
69
|
+
return { type: 'table', rows };
|
|
70
|
+
}
|
|
71
|
+
/**
|
|
72
|
+
* Convert HTML snippet into ContentBlocks (structural convert path).
|
|
73
|
+
*/
|
|
74
|
+
export function htmlToBlocks(html) {
|
|
75
|
+
const normalized = html
|
|
76
|
+
.replace(/\r\n/g, '\n')
|
|
77
|
+
.replace(/<script[\s\S]*?<\/script>/gi, '')
|
|
78
|
+
.replace(/<style[\s\S]*?<\/style>/gi, '');
|
|
79
|
+
const blocks = [];
|
|
80
|
+
const re = /<(h([1-6]))\b[^>]*>([\s\S]*?)<\/\1>|<p\b[^>]*>([\s\S]*?)<\/p>|<li\b[^>]*>([\s\S]*?)<\/li>|<table\b[^>]*>([\s\S]*?)<\/table>|<hr\s*\/?>|<pre\b[^>]*>([\s\S]*?)<\/pre>|<ul\b[^>]*>([\s\S]*?)<\/ul>|<ol\b[^>]*>([\s\S]*?)<\/ol>/gi;
|
|
81
|
+
let m;
|
|
82
|
+
let last = 0;
|
|
83
|
+
const pushLoose = (chunk) => {
|
|
84
|
+
const plain = decodeEntities(chunk.replace(/<[^>]+>/g, '')).trim();
|
|
85
|
+
if (plain)
|
|
86
|
+
blocks.push({ type: 'paragraph', runs: [{ text: plain }] });
|
|
87
|
+
};
|
|
88
|
+
while ((m = re.exec(normalized)) !== null) {
|
|
89
|
+
if (m.index > last)
|
|
90
|
+
pushLoose(normalized.slice(last, m.index));
|
|
91
|
+
last = m.index + m[0].length;
|
|
92
|
+
if (m[2]) {
|
|
93
|
+
const level = Number(m[2]);
|
|
94
|
+
blocks.push({ type: 'heading', level, runs: parseInline(m[3]) });
|
|
95
|
+
}
|
|
96
|
+
else if (m[4] != null && m[0].toLowerCase().startsWith('<p')) {
|
|
97
|
+
const runs = parseInline(m[4]);
|
|
98
|
+
if (runsToText(runs).trim())
|
|
99
|
+
blocks.push({ type: 'paragraph', runs });
|
|
100
|
+
}
|
|
101
|
+
else if (m[5] != null && /<li/i.test(m[0])) {
|
|
102
|
+
blocks.push({ type: 'list_item', level: 0, runs: parseInline(m[5]) });
|
|
103
|
+
}
|
|
104
|
+
else if (m[6] != null && /<table/i.test(m[0])) {
|
|
105
|
+
const table = parseTableHtml(m[0]);
|
|
106
|
+
if (table.type === 'table' && table.rows.length)
|
|
107
|
+
blocks.push(table);
|
|
108
|
+
}
|
|
109
|
+
else if (/^<hr/i.test(m[0])) {
|
|
110
|
+
blocks.push({ type: 'thematic_break' });
|
|
111
|
+
}
|
|
112
|
+
else if (m[7] != null && /<pre/i.test(m[0])) {
|
|
113
|
+
blocks.push({
|
|
114
|
+
type: 'code',
|
|
115
|
+
text: decodeEntities(m[7].replace(/<[^>]+>/g, '')),
|
|
116
|
+
});
|
|
117
|
+
}
|
|
118
|
+
else if (m[8] != null && /<ul/i.test(m[0])) {
|
|
119
|
+
const liRe = /<li\b[^>]*>([\s\S]*?)<\/li>/gi;
|
|
120
|
+
let li;
|
|
121
|
+
while ((li = liRe.exec(m[8])) !== null) {
|
|
122
|
+
blocks.push({ type: 'list_item', level: 0, runs: parseInline(li[1]) });
|
|
123
|
+
}
|
|
124
|
+
}
|
|
125
|
+
else if (m[9] != null && /<ol/i.test(m[0])) {
|
|
126
|
+
const liRe = /<li\b[^>]*>([\s\S]*?)<\/li>/gi;
|
|
127
|
+
let li;
|
|
128
|
+
while ((li = liRe.exec(m[9])) !== null) {
|
|
129
|
+
blocks.push({
|
|
130
|
+
type: 'list_item',
|
|
131
|
+
level: 0,
|
|
132
|
+
ordered: true,
|
|
133
|
+
runs: parseInline(li[1]),
|
|
134
|
+
});
|
|
135
|
+
}
|
|
136
|
+
}
|
|
137
|
+
}
|
|
138
|
+
if (last < normalized.length)
|
|
139
|
+
pushLoose(normalized.slice(last));
|
|
140
|
+
return blocks;
|
|
141
|
+
}
|
|
142
|
+
function runsToText(runs) {
|
|
143
|
+
return runs.map((r) => r.text).join('');
|
|
144
|
+
}
|
|
145
|
+
export function htmlToStructured(html) {
|
|
146
|
+
const blocks = htmlToBlocks(html);
|
|
147
|
+
const paragraphs = flattenBlocksToParagraphs(blocks);
|
|
148
|
+
return {
|
|
149
|
+
format: 'html',
|
|
150
|
+
blocks,
|
|
151
|
+
paragraphs,
|
|
152
|
+
text: paragraphs.join('\n\n'),
|
|
153
|
+
};
|
|
154
|
+
}
|
|
155
|
+
/** Lightweight MD → blocks (headings, lists, GFM tables, paragraphs). */
|
|
156
|
+
export function markdownToBlocks(md) {
|
|
157
|
+
const lines = md.replace(/\r\n/g, '\n').split('\n');
|
|
158
|
+
const blocks = [];
|
|
159
|
+
let i = 0;
|
|
160
|
+
while (i < lines.length) {
|
|
161
|
+
const line = lines[i];
|
|
162
|
+
if (/^\s*$/.test(line)) {
|
|
163
|
+
i++;
|
|
164
|
+
continue;
|
|
165
|
+
}
|
|
166
|
+
const hm = /^(#{1,6})\s+(.*)$/.exec(line);
|
|
167
|
+
if (hm) {
|
|
168
|
+
blocks.push({
|
|
169
|
+
type: 'heading',
|
|
170
|
+
level: hm[1].length,
|
|
171
|
+
runs: [{ text: hm[2] }],
|
|
172
|
+
});
|
|
173
|
+
i++;
|
|
174
|
+
continue;
|
|
175
|
+
}
|
|
176
|
+
if (/^---+$/.test(line.trim())) {
|
|
177
|
+
blocks.push({ type: 'thematic_break' });
|
|
178
|
+
i++;
|
|
179
|
+
continue;
|
|
180
|
+
}
|
|
181
|
+
if (/^\s*[-*+]\s+/.test(line)) {
|
|
182
|
+
while (i < lines.length && /^\s*[-*+]\s+/.test(lines[i])) {
|
|
183
|
+
const text = lines[i].replace(/^\s*[-*+]\s+/, '');
|
|
184
|
+
blocks.push({ type: 'list_item', level: 0, runs: [{ text }] });
|
|
185
|
+
i++;
|
|
186
|
+
}
|
|
187
|
+
continue;
|
|
188
|
+
}
|
|
189
|
+
if (/^\s*\d+\.\s+/.test(line)) {
|
|
190
|
+
while (i < lines.length && /^\s*\d+\.\s+/.test(lines[i])) {
|
|
191
|
+
const text = lines[i].replace(/^\s*\d+\.\s+/, '');
|
|
192
|
+
blocks.push({
|
|
193
|
+
type: 'list_item',
|
|
194
|
+
level: 0,
|
|
195
|
+
ordered: true,
|
|
196
|
+
runs: [{ text }],
|
|
197
|
+
});
|
|
198
|
+
i++;
|
|
199
|
+
}
|
|
200
|
+
continue;
|
|
201
|
+
}
|
|
202
|
+
if (/^\s*\|/.test(line) && i + 1 < lines.length && /^\s*\|?\s*:?-+:?/.test(lines[i + 1])) {
|
|
203
|
+
const rows = [];
|
|
204
|
+
while (i < lines.length && /^\s*\|/.test(lines[i])) {
|
|
205
|
+
if (/^\s*\|?\s*:?-+:?/.test(lines[i])) {
|
|
206
|
+
i++;
|
|
207
|
+
continue;
|
|
208
|
+
}
|
|
209
|
+
const cells = lines[i]
|
|
210
|
+
.replace(/^\s*\|/, '')
|
|
211
|
+
.replace(/\|\s*$/, '')
|
|
212
|
+
.split('|')
|
|
213
|
+
.map((c) => c.trim());
|
|
214
|
+
rows.push(cells);
|
|
215
|
+
i++;
|
|
216
|
+
}
|
|
217
|
+
if (rows.length)
|
|
218
|
+
blocks.push({ type: 'table', rows });
|
|
219
|
+
continue;
|
|
220
|
+
}
|
|
221
|
+
if (/^```/.test(line)) {
|
|
222
|
+
const lang = line.replace(/^```/, '').trim();
|
|
223
|
+
i++;
|
|
224
|
+
const buf = [];
|
|
225
|
+
while (i < lines.length && !/^```/.test(lines[i])) {
|
|
226
|
+
buf.push(lines[i]);
|
|
227
|
+
i++;
|
|
228
|
+
}
|
|
229
|
+
if (i < lines.length)
|
|
230
|
+
i++;
|
|
231
|
+
blocks.push({ type: 'code', text: buf.join('\n'), language: lang || undefined });
|
|
232
|
+
continue;
|
|
233
|
+
}
|
|
234
|
+
const para = [line];
|
|
235
|
+
i++;
|
|
236
|
+
while (i < lines.length && !/^\s*$/.test(lines[i]) && !/^(#{1,6})\s+|^\s*[-*+]\s+|^\s*\||^```/.test(lines[i])) {
|
|
237
|
+
para.push(lines[i]);
|
|
238
|
+
i++;
|
|
239
|
+
}
|
|
240
|
+
blocks.push({ type: 'paragraph', runs: [{ text: para.join(' ') }] });
|
|
241
|
+
}
|
|
242
|
+
return blocks;
|
|
243
|
+
}
|
|
244
|
+
//# sourceMappingURL=html-blocks.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"html-blocks.js","sourceRoot":"","sources":["../../src/parse/html-blocks.ts"],"names":[],"mappings":"AACA,OAAO,EAAE,yBAAyB,EAAE,MAAM,sBAAsB,CAAC;AAEjE,SAAS,cAAc,CAAC,CAAS;IAC/B,OAAO,CAAC;SACL,OAAO,CAAC,UAAU,EAAE,GAAG,CAAC;SACxB,OAAO,CAAC,QAAQ,EAAE,GAAG,CAAC;SACtB,OAAO,CAAC,QAAQ,EAAE,GAAG,CAAC;SACtB,OAAO,CAAC,UAAU,EAAE,GAAG,CAAC;SACxB,OAAO,CAAC,UAAU,EAAE,GAAG,CAAC;SACxB,OAAO,CAAC,SAAS,EAAE,GAAG,CAAC,CAAC;AAC7B,CAAC;AAED,SAAS,WAAW,CAAC,IAAY;IAC/B,MAAM,IAAI,GAAgB,EAAE,CAAC;IAC7B,MAAM,EAAE,GACN,6FAA6F,CAAC;IAChG,IAAI,CAAyB,CAAC;IAC9B,OAAO,CAAC,CAAC,GAAG,EAAE,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACpC,IAAI,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;YACT,MAAM,KAAK,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;YACnB,8BAA8B;YAC9B,MAAM,MAAM,GAAG,qCAAqC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;YACjE,IAAI,MAAM,EAAE,CAAC;gBACX,MAAM,MAAM,GAAG,cAAc,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,EAAE,MAAM,CAAC,KAAK,CAAC,CAAC,OAAO,CAAC,UAAU,EAAE,EAAE,CAAC,CAAC,CAAC;gBACpF,IAAI,MAAM;oBAAE,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,IAAI,EAAE,CAAC,CAAC;gBACpD,MAAM,GAAG,GAAG,cAAc,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,UAAU,EAAE,EAAE,CAAC,CAAC,CAAC;gBAC9D,IAAI,GAAG;oBAAE,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,GAAG,EAAE,IAAI,EAAE,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,CAAC,CAAC;gBAC5D,MAAM,KAAK,GAAG,cAAc,CAAC,KAAK,CAAC,KAAK,CAAC,MAAM,CAAC,KAAK,GAAG,MAAM,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,CAAC,OAAO,CAAC,UAAU,EAAE,EAAE,CAAC,CAAC,CAAC;gBACnG,IAAI,KAAK;oBAAE,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,KAAK,EAAE,IAAI,EAAE,IAAI,EAAE,CAAC,CAAC;YACpD,CAAC;iBAAM,CAAC;gBACN,MAAM,IAAI,GAAG,cAAc,CAAC,KAAK,CAAC,OAAO,CAAC,UAAU,EAAE,EAAE,CAAC,CAAC,CAAC;gBAC3D,IAAI,IAAI;oBAAE,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE,IAAI,EAAE,CAAC,CAAC;YAC5C,CAAC;QACH,CAAC;aAAM,IAAI,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;YAChB,MAAM,IAAI,GAAG,cAAc,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,UAAU,EAAE,EAAE,CAAC,CAAC,CAAC;YAC1D,IAAI,IAAI;gBAAE,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,CAAC,CAAC;QAC9C,CAAC;aAAM,IAAI,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;YAChB,MAAM,CAAC,GAAG,cAAc,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;YAC/B,IAAI,CAAC;gBAAE,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,CAAC,CAAC;QAChC,CAAC;IACH,CAAC;IACD,IAAI,CAAC,IAAI,CAAC,MAAM,EAAE,CAAC;QACjB,MAAM,KAAK,GAAG,cAAc,CAAC,IAAI,CAAC,OAAO,CAAC,UAAU,EAAE,EAAE,CAAC,CAAC,CAAC;QAC3D,IAAI,KAAK;YAAE,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,KAAK,EAAE,CAAC,CAAC;IACxC,CAAC;IACD,OAAO,IAAI,CAAC;AACd,CAAC;AAED,SAAS,cAAc,CAAC,SAAiB;IACvC,MAAM,IAAI,GAAe,EAAE,CAAC;IAC5B,MAAM,IAAI,GAAG,+BAA+B,CAAC;IAC7C,IAAI,EAA0B,CAAC;IAC/B,OAAO,CAAC,EAAE,GAAG,IAAI,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QAC5C,MAAM,KAAK,GAAa,EAAE,CAAC;QAC3B,MAAM,MAAM,GAAG,oCAAoC,CAAC;QACpD,IAAI,CAAyB,CAAC;QAC9B,OAAO,CAAC,CAAC,GAAG,MAAM,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;YACzC,KAAK,CAAC,IAAI,CAAC,cAAc,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,UAAU,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC;QAClE,CAAC;QACD,IAAI,KAAK,CAAC,MAAM;YAAE,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;IACrC,CAAC;IACD,OAAO,EAAE,IAAI,EAAE,OAAO,EAAE,IAAI,EAAE,CAAC;AACjC,CAAC;AAED;;GAEG;AACH,MAAM,UAAU,YAAY,CAAC,IAAY;IACvC,MAAM,UAAU,GAAG,IAAI;SACpB,OAAO,CAAC,OAAO,EAAE,IAAI,CAAC;SACtB,OAAO,CAAC,6BAA6B,EAAE,EAAE,CAAC;SAC1C,OAAO,CAAC,2BAA2B,EAAE,EAAE,CAAC,CAAC;IAE5C,MAAM,MAAM,GAAmB,EAAE,CAAC;IAClC,MAAM,EAAE,GACN,gOAAgO,CAAC;IAEnO,IAAI,CAAyB,CAAC;IAC9B,IAAI,IAAI,GAAG,CAAC,CAAC;IACb,MAAM,SAAS,GAAG,CAAC,KAAa,EAAE,EAAE;QAClC,MAAM,KAAK,GAAG,cAAc,CAAC,KAAK,CAAC,OAAO,CAAC,UAAU,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;QACnE,IAAI,KAAK;YAAE,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,WAAW,EAAE,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,KAAK,EAAE,CAAC,EAAE,CAAC,CAAC;IACzE,CAAC,CAAC;IAEF,OAAO,CAAC,CAAC,GAAG,EAAE,CAAC,IAAI,CAAC,UAAU,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QAC1C,IAAI,CAAC,CAAC,KAAK,GAAG,IAAI;YAAE,SAAS,CAAC,UAAU,CAAC,KAAK,CAAC,IAAI,EAAE,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC;QAC/D,IAAI,GAAG,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC;QAE7B,IAAI,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;YACT,MAAM,KAAK,GAAG,MAAM,CAAC,CAAC,CAAC,CAAC,CAAC,CAA0B,CAAC;YACpD,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,SAAS,EAAE,KAAK,EAAE,IAAI,EAAE,WAAW,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC;QACnE,CAAC;aAAM,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,IAAI,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,WAAW,EAAE,CAAC,UAAU,CAAC,IAAI,CAAC,EAAE,CAAC;YAC/D,MAAM,IAAI,GAAG,WAAW,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;YAC/B,IAAI,UAAU,CAAC,IAAI,CAAC,CAAC,IAAI,EAAE;gBAAE,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,WAAW,EAAE,IAAI,EAAE,CAAC,CAAC;QACxE,CAAC;aAAM,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,IAAI,IAAI,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;YAC7C,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,WAAW,EAAE,KAAK,EAAE,CAAC,EAAE,IAAI,EAAE,WAAW,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC;QACxE,CAAC;aAAM,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,IAAI,IAAI,SAAS,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;YAChD,MAAM,KAAK,GAAG,cAAc,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;YACnC,IAAI,KAAK,CAAC,IAAI,KAAK,OAAO,IAAI,KAAK,CAAC,IAAI,CAAC,MAAM;gBAAE,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QACtE,CAAC;aAAM,IAAI,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;YAC9B,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,gBAAgB,EAAE,CAAC,CAAC;QAC1C,CAAC;aAAM,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,IAAI,IAAI,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;YAC9C,MAAM,CAAC,IAAI,CAAC;gBACV,IAAI,EAAE,MAAM;gBACZ,IAAI,EAAE,cAAc,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,UAAU,EAAE,EAAE,CAAC,CAAC;aACnD,CAAC,CAAC;QACL,CAAC;aAAM,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,IAAI,IAAI,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;YAC7C,MAAM,IAAI,GAAG,+BAA+B,CAAC;YAC7C,IAAI,EAA0B,CAAC;YAC/B,OAAO,CAAC,EAAE,GAAG,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;gBACvC,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,WAAW,EAAE,KAAK,EAAE,CAAC,EAAE,IAAI,EAAE,WAAW,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC;YACzE,CAAC;QACH,CAAC;aAAM,IAAI,CAAC,CAAC,CAAC,CAAC,IAAI,IAAI,IAAI,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;YAC7C,MAAM,IAAI,GAAG,+BAA+B,CAAC;YAC7C,IAAI,EAA0B,CAAC;YAC/B,OAAO,CAAC,EAAE,GAAG,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;gBACvC,MAAM,CAAC,IAAI,CAAC;oBACV,IAAI,EAAE,WAAW;oBACjB,KAAK,EAAE,CAAC;oBACR,OAAO,EAAE,IAAI;oBACb,IAAI,EAAE,WAAW,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC;iBACzB,CAAC,CAAC;YACL,CAAC;QACH,CAAC;IACH,CAAC;IACD,IAAI,IAAI,GAAG,UAAU,CAAC,MAAM;QAAE,SAAS,CAAC,UAAU,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC;IAEhE,OAAO,MAAM,CAAC;AAChB,CAAC;AAED,SAAS,UAAU,CAAC,IAAiB;IACnC,OAAO,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;AAC1C,CAAC;AAED,MAAM,UAAU,gBAAgB,CAAC,IAAY;IAM3C,MAAM,MAAM,GAAG,YAAY,CAAC,IAAI,CAAC,CAAC;IAClC,MAAM,UAAU,GAAG,yBAAyB,CAAC,MAAM,CAAC,CAAC;IACrD,OAAO;QACL,MAAM,EAAE,MAAM;QACd,MAAM;QACN,UAAU;QACV,IAAI,EAAE,UAAU,CAAC,IAAI,CAAC,MAAM,CAAC;KAC9B,CAAC;AACJ,CAAC;AAED,yEAAyE;AACzE,MAAM,UAAU,gBAAgB,CAAC,EAAU;IACzC,MAAM,KAAK,GAAG,EAAE,CAAC,OAAO,CAAC,OAAO,EAAE,IAAI,CAAC,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;IACpD,MAAM,MAAM,GAAmB,EAAE,CAAC;IAClC,IAAI,CAAC,GAAG,CAAC,CAAC;IACV,OAAO,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC;QACxB,MAAM,IAAI,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC;QACtB,IAAI,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;YACvB,CAAC,EAAE,CAAC;YACJ,SAAS;QACX,CAAC;QACD,MAAM,EAAE,GAAG,mBAAmB,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QAC1C,IAAI,EAAE,EAAE,CAAC;YACP,MAAM,CAAC,IAAI,CAAC;gBACV,IAAI,EAAE,SAAS;gBACf,KAAK,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,MAA+B;gBAC5C,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;aACxB,CAAC,CAAC;YACH,CAAC,EAAE,CAAC;YACJ,SAAS;QACX,CAAC;QACD,IAAI,QAAQ,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,EAAE,CAAC,EAAE,CAAC;YAC/B,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,gBAAgB,EAAE,CAAC,CAAC;YACxC,CAAC,EAAE,CAAC;YACJ,SAAS;QACX,CAAC;QACD,IAAI,cAAc,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;YAC9B,OAAO,CAAC,GAAG,KAAK,CAAC,MAAM,IAAI,cAAc,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;gBACzD,MAAM,IAAI,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,cAAc,EAAE,EAAE,CAAC,CAAC;gBAClD,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,WAAW,EAAE,KAAK,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC,EAAE,CAAC,CAAC;gBAC/D,CAAC,EAAE,CAAC;YACN,CAAC;YACD,SAAS;QACX,CAAC;QACD,IAAI,cAAc,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;YAC9B,OAAO,CAAC,GAAG,KAAK,CAAC,MAAM,IAAI,cAAc,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;gBACzD,MAAM,IAAI,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,cAAc,EAAE,EAAE,CAAC,CAAC;gBAClD,MAAM,CAAC,IAAI,CAAC;oBACV,IAAI,EAAE,WAAW;oBACjB,KAAK,EAAE,CAAC;oBACR,OAAO,EAAE,IAAI;oBACb,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,CAAC;iBACjB,CAAC,CAAC;gBACH,CAAC,EAAE,CAAC;YACN,CAAC;YACD,SAAS;QACX,CAAC;QACD,IAAI,QAAQ,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,GAAG,KAAK,CAAC,MAAM,IAAI,kBAAkB,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC;YACzF,MAAM,IAAI,GAAe,EAAE,CAAC;YAC5B,OAAO,CAAC,GAAG,KAAK,CAAC,MAAM,IAAI,QAAQ,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;gBACnD,IAAI,kBAAkB,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;oBACtC,CAAC,EAAE,CAAC;oBACJ,SAAS;gBACX,CAAC;gBACD,MAAM,KAAK,GAAG,KAAK,CAAC,CAAC,CAAC;qBACnB,OAAO,CAAC,QAAQ,EAAE,EAAE,CAAC;qBACrB,OAAO,CAAC,QAAQ,EAAE,EAAE,CAAC;qBACrB,KAAK,CAAC,GAAG,CAAC;qBACV,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC;gBACxB,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;gBACjB,CAAC,EAAE,CAAC;YACN,CAAC;YACD,IAAI,IAAI,CAAC,MAAM;gBAAE,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,OAAO,EAAE,IAAI,EAAE,CAAC,CAAC;YACtD,SAAS;QACX,CAAC;QACD,IAAI,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC;YACtB,MAAM,IAAI,GAAG,IAAI,CAAC,OAAO,CAAC,MAAM,EAAE,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC;YAC7C,CAAC,EAAE,CAAC;YACJ,MAAM,GAAG,GAAa,EAAE,CAAC;YACzB,OAAO,CAAC,GAAG,KAAK,CAAC,MAAM,IAAI,CAAC,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;gBAClD,GAAG,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC;gBACnB,CAAC,EAAE,CAAC;YACN,CAAC;YACD,IAAI,CAAC,GAAG,KAAK,CAAC,MAAM;gBAAE,CAAC,EAAE,CAAC;YAC1B,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE,QAAQ,EAAE,IAAI,IAAI,SAAS,EAAE,CAAC,CAAC;YACjF,SAAS;QACX,CAAC;QACD,MAAM,IAAI,GAAa,CAAC,IAAI,CAAC,CAAC;QAC9B,CAAC,EAAE,CAAC;QACJ,OAAO,CAAC,GAAG,KAAK,CAAC,MAAM,IAAI,CAAC,OAAO,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,uCAAuC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,EAAE,CAAC;YAC9G,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC;YACpB,CAAC,EAAE,CAAC;QACN,CAAC;QACD,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,WAAW,EAAE,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,EAAE,CAAC,CAAC;IACvE,CAAC;IACD,OAAO,MAAM,CAAC;AAChB,CAAC"}
|
package/dist/parse/pptx.d.ts
CHANGED
|
@@ -1,7 +1,6 @@
|
|
|
1
1
|
import type { PptxStructured } from '../types.js';
|
|
2
2
|
export declare function readPptxStructured(filePath: string): Promise<PptxStructured>;
|
|
3
3
|
export declare function readPptxStructuredFromBuffer(buf: Buffer): Promise<PptxStructured>;
|
|
4
|
-
/** Flatten slides to paragraphs (for text / convert emit). */
|
|
5
4
|
export declare function pptxToParagraphs(structured: PptxStructured): string[];
|
|
6
5
|
export declare function pptxToText(structured: PptxStructured): string;
|
|
7
6
|
//# sourceMappingURL=pptx.d.ts.map
|
package/dist/parse/pptx.d.ts.map
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"pptx.d.ts","sourceRoot":"","sources":["../../src/parse/pptx.ts"],"names":[],"mappings":"AAGA,OAAO,KAAK,
|
|
1
|
+
{"version":3,"file":"pptx.d.ts","sourceRoot":"","sources":["../../src/parse/pptx.ts"],"names":[],"mappings":"AAGA,OAAO,KAAK,EAA2B,cAAc,EAAE,MAAM,aAAa,CAAC;AA2I3E,wBAAsB,kBAAkB,CAAC,QAAQ,EAAE,MAAM,GAAG,OAAO,CAAC,cAAc,CAAC,CAQlF;AAED,wBAAsB,4BAA4B,CAChD,GAAG,EAAE,MAAM,GACV,OAAO,CAAC,cAAc,CAAC,CAQzB;AAED,wBAAgB,gBAAgB,CAAC,UAAU,EAAE,cAAc,GAAG,MAAM,EAAE,CAarE;AAED,wBAAgB,UAAU,CAAC,UAAU,EAAE,cAAc,GAAG,MAAM,CAQ7D"}
|
package/dist/parse/pptx.js
CHANGED
|
@@ -1,6 +1,7 @@
|
|
|
1
1
|
import { readFile } from 'node:fs/promises';
|
|
2
2
|
import JSZip from 'jszip';
|
|
3
3
|
import { DocumentError, wrapError } from '../utils/errors.js';
|
|
4
|
+
import { flattenBlocksToParagraphs } from '../convert/blocks.js';
|
|
4
5
|
function decodeXmlEntities(s) {
|
|
5
6
|
return s
|
|
6
7
|
.replace(/</g, '<')
|
|
@@ -9,33 +10,96 @@ function decodeXmlEntities(s) {
|
|
|
9
10
|
.replace(/'/g, "'")
|
|
10
11
|
.replace(/&/g, '&');
|
|
11
12
|
}
|
|
12
|
-
|
|
13
|
-
|
|
13
|
+
function parsePptxRuns(paraXml) {
|
|
14
|
+
const runs = [];
|
|
15
|
+
const runRe = /<a:r\b[^>]*>([\s\S]*?)<\/a:r>/g;
|
|
16
|
+
let rm;
|
|
17
|
+
while ((rm = runRe.exec(paraXml)) !== null) {
|
|
18
|
+
const inner = rm[1];
|
|
19
|
+
const bold = /b="1"/.test(inner) || /<a:b\b/.test(inner);
|
|
20
|
+
const italic = /i="1"/.test(inner) || /<a:i\b/.test(inner);
|
|
21
|
+
const tm = /<a:t(?:\s[^>]*)?>([^<]*)<\/a:t>/.exec(inner);
|
|
22
|
+
if (!tm)
|
|
23
|
+
continue;
|
|
24
|
+
const text = decodeXmlEntities(tm[1]);
|
|
25
|
+
if (!text)
|
|
26
|
+
continue;
|
|
27
|
+
runs.push({
|
|
28
|
+
text,
|
|
29
|
+
...(bold ? { bold: true } : {}),
|
|
30
|
+
...(italic ? { italic: true } : {}),
|
|
31
|
+
});
|
|
32
|
+
}
|
|
33
|
+
if (!runs.length) {
|
|
34
|
+
const tRe = /<a:t(?:\s[^>]*)?>([^<]*)<\/a:t>/g;
|
|
35
|
+
let tm;
|
|
36
|
+
while ((tm = tRe.exec(paraXml)) !== null) {
|
|
37
|
+
const text = decodeXmlEntities(tm[1]);
|
|
38
|
+
if (text)
|
|
39
|
+
runs.push({ text });
|
|
40
|
+
}
|
|
41
|
+
}
|
|
42
|
+
return runs;
|
|
43
|
+
}
|
|
44
|
+
function extractSlideTables(xml) {
|
|
45
|
+
const blocks = [];
|
|
46
|
+
const tblRe = /<a:tbl\b[\s\S]*?<\/a:tbl>/g;
|
|
47
|
+
let tm;
|
|
48
|
+
while ((tm = tblRe.exec(xml)) !== null) {
|
|
49
|
+
const rows = [];
|
|
50
|
+
const trRe = /<a:tr\b[\s\S]*?<\/a:tr>/g;
|
|
51
|
+
let tr;
|
|
52
|
+
while ((tr = trRe.exec(tm[0])) !== null) {
|
|
53
|
+
const cells = [];
|
|
54
|
+
const tcRe = /<a:tc\b[\s\S]*?<\/a:tc>/g;
|
|
55
|
+
let tc;
|
|
56
|
+
while ((tc = tcRe.exec(tr[0])) !== null) {
|
|
57
|
+
const texts = [];
|
|
58
|
+
const tRe = /<a:t(?:\s[^>]*)?>([^<]*)<\/a:t>/g;
|
|
59
|
+
let t;
|
|
60
|
+
while ((t = tRe.exec(tc[0])) !== null) {
|
|
61
|
+
texts.push(decodeXmlEntities(t[1]));
|
|
62
|
+
}
|
|
63
|
+
cells.push(texts.join('').trim());
|
|
64
|
+
}
|
|
65
|
+
if (cells.length)
|
|
66
|
+
rows.push(cells);
|
|
67
|
+
}
|
|
68
|
+
if (rows.length)
|
|
69
|
+
blocks.push({ type: 'table', rows });
|
|
70
|
+
}
|
|
71
|
+
return blocks;
|
|
72
|
+
}
|
|
73
|
+
function extractSlideBlocks(xml, slideIndex) {
|
|
14
74
|
const texts = [];
|
|
75
|
+
const blocks = [
|
|
76
|
+
{
|
|
77
|
+
type: 'heading',
|
|
78
|
+
level: 2,
|
|
79
|
+
runs: [{ text: `Slide ${slideIndex + 1}` }],
|
|
80
|
+
},
|
|
81
|
+
];
|
|
82
|
+
// Remove tables from paragraph scan to avoid duplicating cell text as paras
|
|
83
|
+
const withoutTables = xml.replace(/<a:tbl\b[\s\S]*?<\/a:tbl>/g, '');
|
|
15
84
|
const paraRegex = /<a:p[\s\S]*?<\/a:p>/g;
|
|
16
85
|
let pm;
|
|
17
|
-
|
|
18
|
-
|
|
19
|
-
const
|
|
20
|
-
|
|
21
|
-
|
|
22
|
-
|
|
86
|
+
let paraIdx = 0;
|
|
87
|
+
while ((pm = paraRegex.exec(withoutTables)) !== null) {
|
|
88
|
+
const runs = parsePptxRuns(pm[0]);
|
|
89
|
+
const line = runs.map((r) => r.text).join('').trim();
|
|
90
|
+
if (!line)
|
|
91
|
+
continue;
|
|
92
|
+
texts.push(line);
|
|
93
|
+
if (paraIdx === 0) {
|
|
94
|
+
blocks.push({ type: 'heading', level: 3, runs: runs.length ? runs : [{ text: line }] });
|
|
23
95
|
}
|
|
24
|
-
|
|
25
|
-
|
|
26
|
-
texts.push(line);
|
|
27
|
-
}
|
|
28
|
-
// Fallback: any a:t outside well-formed a:p (rare)
|
|
29
|
-
if (texts.length === 0) {
|
|
30
|
-
const tRegex = /<a:t(?:\s[^>]*)?>([^<]*)<\/a:t>/g;
|
|
31
|
-
let tm;
|
|
32
|
-
while ((tm = tRegex.exec(xml)) !== null) {
|
|
33
|
-
const line = decodeXmlEntities(tm[1]).trim();
|
|
34
|
-
if (line)
|
|
35
|
-
texts.push(line);
|
|
96
|
+
else {
|
|
97
|
+
blocks.push({ type: 'paragraph', runs: runs.length ? runs : [{ text: line }] });
|
|
36
98
|
}
|
|
99
|
+
paraIdx++;
|
|
37
100
|
}
|
|
38
|
-
|
|
101
|
+
blocks.push(...extractSlideTables(xml));
|
|
102
|
+
return { texts, blocks };
|
|
39
103
|
}
|
|
40
104
|
function listSlidePaths(zip) {
|
|
41
105
|
return Object.keys(zip.files)
|
|
@@ -51,14 +115,19 @@ async function readPptxFromZip(zip) {
|
|
|
51
115
|
throw new DocumentError('PARSE_FAILED', 'Invalid pptx: missing ppt/presentation.xml / slides');
|
|
52
116
|
}
|
|
53
117
|
const slides = [];
|
|
118
|
+
const blocks = [];
|
|
119
|
+
let i = 0;
|
|
54
120
|
for (const path of listSlidePaths(zip)) {
|
|
55
121
|
const file = zip.file(path);
|
|
56
122
|
if (!file)
|
|
57
123
|
continue;
|
|
58
124
|
const xml = await file.async('string');
|
|
59
|
-
|
|
125
|
+
const { texts, blocks: slideBlocks } = extractSlideBlocks(xml, i);
|
|
126
|
+
slides.push(texts);
|
|
127
|
+
blocks.push(...slideBlocks);
|
|
128
|
+
i++;
|
|
60
129
|
}
|
|
61
|
-
return { format: 'pptx', slides };
|
|
130
|
+
return { format: 'pptx', slides, blocks };
|
|
62
131
|
}
|
|
63
132
|
export async function readPptxStructured(filePath) {
|
|
64
133
|
try {
|
|
@@ -82,8 +151,10 @@ export async function readPptxStructuredFromBuffer(buf) {
|
|
|
82
151
|
throw wrapError('PARSE_FAILED', 'Failed to parse pptx buffer', err);
|
|
83
152
|
}
|
|
84
153
|
}
|
|
85
|
-
/** Flatten slides to paragraphs (for text / convert emit). */
|
|
86
154
|
export function pptxToParagraphs(structured) {
|
|
155
|
+
if (structured.blocks?.length) {
|
|
156
|
+
return flattenBlocksToParagraphs(structured.blocks);
|
|
157
|
+
}
|
|
87
158
|
const out = [];
|
|
88
159
|
structured.slides.forEach((texts, i) => {
|
|
89
160
|
if (texts.length === 0) {
|
package/dist/parse/pptx.js.map
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"pptx.js","sourceRoot":"","sources":["../../src/parse/pptx.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AAC5C,OAAO,KAAK,MAAM,OAAO,CAAC;AAC1B,OAAO,EAAE,aAAa,EAAE,SAAS,EAAE,MAAM,oBAAoB,CAAC;
|
|
1
|
+
{"version":3,"file":"pptx.js","sourceRoot":"","sources":["../../src/parse/pptx.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,kBAAkB,CAAC;AAC5C,OAAO,KAAK,MAAM,OAAO,CAAC;AAC1B,OAAO,EAAE,aAAa,EAAE,SAAS,EAAE,MAAM,oBAAoB,CAAC;AAE9D,OAAO,EAAE,yBAAyB,EAAE,MAAM,sBAAsB,CAAC;AAEjE,SAAS,iBAAiB,CAAC,CAAS;IAClC,OAAO,CAAC;SACL,OAAO,CAAC,OAAO,EAAE,GAAG,CAAC;SACrB,OAAO,CAAC,OAAO,EAAE,GAAG,CAAC;SACrB,OAAO,CAAC,SAAS,EAAE,GAAG,CAAC;SACvB,OAAO,CAAC,SAAS,EAAE,GAAG,CAAC;SACvB,OAAO,CAAC,QAAQ,EAAE,GAAG,CAAC,CAAC;AAC5B,CAAC;AAED,SAAS,aAAa,CAAC,OAAe;IACpC,MAAM,IAAI,GAAgB,EAAE,CAAC;IAC7B,MAAM,KAAK,GAAG,gCAAgC,CAAC;IAC/C,IAAI,EAA0B,CAAC;IAC/B,OAAO,CAAC,EAAE,GAAG,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QAC3C,MAAM,KAAK,GAAG,EAAE,CAAC,CAAC,CAAC,CAAC;QACpB,MAAM,IAAI,GAAG,OAAO,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,QAAQ,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QACzD,MAAM,MAAM,GAAG,OAAO,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,QAAQ,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QAC3D,MAAM,EAAE,GAAG,iCAAiC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QACzD,IAAI,CAAC,EAAE;YAAE,SAAS;QAClB,MAAM,IAAI,GAAG,iBAAiB,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC;QACtC,IAAI,CAAC,IAAI;YAAE,SAAS;QACpB,IAAI,CAAC,IAAI,CAAC;YACR,IAAI;YACJ,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;YAC/B,GAAG,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,MAAM,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;SACpC,CAAC,CAAC;IACL,CAAC;IACD,IAAI,CAAC,IAAI,CAAC,MAAM,EAAE,CAAC;QACjB,MAAM,GAAG,GAAG,kCAAkC,CAAC;QAC/C,IAAI,EAA0B,CAAC;QAC/B,OAAO,CAAC,EAAE,GAAG,GAAG,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;YACzC,MAAM,IAAI,GAAG,iBAAiB,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC;YACtC,IAAI,IAAI;gBAAE,IAAI,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,CAAC,CAAC;QAChC,CAAC;IACH,CAAC;IACD,OAAO,IAAI,CAAC;AACd,CAAC;AAED,SAAS,kBAAkB,CAAC,GAAW;IACrC,MAAM,MAAM,GAAmB,EAAE,CAAC;IAClC,MAAM,KAAK,GAAG,4BAA4B,CAAC;IAC3C,IAAI,EAA0B,CAAC;IAC/B,OAAO,CAAC,EAAE,GAAG,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACvC,MAAM,IAAI,GAAe,EAAE,CAAC;QAC5B,MAAM,IAAI,GAAG,0BAA0B,CAAC;QACxC,IAAI,EAA0B,CAAC;QAC/B,OAAO,CAAC,EAAE,GAAG,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;YACxC,MAAM,KAAK,GAAa,EAAE,CAAC;YAC3B,MAAM,IAAI,GAAG,0BAA0B,CAAC;YACxC,IAAI,EAA0B,CAAC;YAC/B,OAAO,CAAC,EAAE,GAAG,IAAI,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;gBACxC,MAAM,KAAK,GAAa,EAAE,CAAC;gBAC3B,MAAM,GAAG,GAAG,kCAAkC,CAAC;gBAC/C,IAAI,CAAyB,CAAC;gBAC9B,OAAO,CAAC,CAAC,GAAG,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;oBACtC,KAAK,CAAC,IAAI,CAAC,iBAAiB,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;gBACtC,CAAC;gBACD,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC;YACpC,CAAC;YACD,IAAI,KAAK,CAAC,MAAM;gBAAE,IAAI,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QACrC,CAAC;QACD,IAAI,IAAI,CAAC,MAAM;YAAE,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,OAAO,EAAE,IAAI,EAAE,CAAC,CAAC;IACxD,CAAC;IACD,OAAO,MAAM,CAAC;AAChB,CAAC;AAED,SAAS,kBAAkB,CAAC,GAAW,EAAE,UAAkB;IAIzD,MAAM,KAAK,GAAa,EAAE,CAAC;IAC3B,MAAM,MAAM,GAAmB;QAC7B;YACE,IAAI,EAAE,SAAS;YACf,KAAK,EAAE,CAAC;YACR,IAAI,EAAE,CAAC,EAAE,IAAI,EAAE,SAAS,UAAU,GAAG,CAAC,EAAE,EAAE,CAAC;SAC5C;KACF,CAAC;IAEF,4EAA4E;IAC5E,MAAM,aAAa,GAAG,GAAG,CAAC,OAAO,CAAC,4BAA4B,EAAE,EAAE,CAAC,CAAC;IAEpE,MAAM,SAAS,GAAG,sBAAsB,CAAC;IACzC,IAAI,EAA0B,CAAC;IAC/B,IAAI,OAAO,GAAG,CAAC,CAAC;IAChB,OAAO,CAAC,EAAE,GAAG,SAAS,CAAC,IAAI,CAAC,aAAa,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACrD,MAAM,IAAI,GAAG,aAAa,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,CAAC;QAClC,MAAM,IAAI,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC,IAAI,EAAE,CAAC;QACrD,IAAI,CAAC,IAAI;YAAE,SAAS;QACpB,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QACjB,IAAI,OAAO,KAAK,CAAC,EAAE,CAAC;YAClB,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,SAAS,EAAE,KAAK,EAAE,CAAC,EAAE,IAAI,EAAE,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE,CAAC,EAAE,CAAC,CAAC;QAC1F,CAAC;aAAM,CAAC;YACN,MAAM,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,WAAW,EAAE,IAAI,EAAE,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,IAAI,EAAE,CAAC,EAAE,CAAC,CAAC;QAClF,CAAC;QACD,OAAO,EAAE,CAAC;IACZ,CAAC;IAED,MAAM,CAAC,IAAI,CAAC,GAAG,kBAAkB,CAAC,GAAG,CAAC,CAAC,CAAC;IACxC,OAAO,EAAE,KAAK,EAAE,MAAM,EAAE,CAAC;AAC3B,CAAC;AAED,SAAS,cAAc,CAAC,GAAU;IAChC,OAAO,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,KAAK,CAAC;SAC1B,MAAM,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,+BAA+B,CAAC,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC;SAC3E,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE;QACb,MAAM,EAAE,GAAG,MAAM,CAAC,aAAa,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC;QACnD,MAAM,EAAE,GAAG,MAAM,CAAC,aAAa,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC;QACnD,OAAO,EAAE,GAAG,EAAE,CAAC;IACjB,CAAC,CAAC,CAAC;AACP,CAAC;AAED,KAAK,UAAU,eAAe,CAAC,GAAU;IACvC,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,sBAAsB,CAAC,IAAI,cAAc,CAAC,GAAG,CAAC,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QAC1E,MAAM,IAAI,aAAa,CACrB,cAAc,EACd,qDAAqD,CACtD,CAAC;IACJ,CAAC;IAED,MAAM,MAAM,GAAe,EAAE,CAAC;IAC9B,MAAM,MAAM,GAAmB,EAAE,CAAC;IAClC,IAAI,CAAC,GAAG,CAAC,CAAC;IACV,KAAK,MAAM,IAAI,IAAI,cAAc,CAAC,GAAG,CAAC,EAAE,CAAC;QACvC,MAAM,IAAI,GAAG,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QAC5B,IAAI,CAAC,IAAI;YAAE,SAAS;QACpB,MAAM,GAAG,GAAG,MAAM,IAAI,CAAC,KAAK,CAAC,QAAQ,CAAC,CAAC;QACvC,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,WAAW,EAAE,GAAG,kBAAkB,CAAC,GAAG,EAAE,CAAC,CAAC,CAAC;QAClE,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QACnB,MAAM,CAAC,IAAI,CAAC,GAAG,WAAW,CAAC,CAAC;QAC5B,CAAC,EAAE,CAAC;IACN,CAAC;IAED,OAAO,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,EAAE,CAAC;AAC5C,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,kBAAkB,CAAC,QAAgB;IACvD,IAAI,CAAC;QACH,MAAM,GAAG,GAAG,MAAM,QAAQ,CAAC,QAAQ,CAAC,CAAC;QACrC,OAAO,4BAA4B,CAAC,GAAG,CAAC,CAAC;IAC3C,CAAC;IAAC,OAAO,GAAG,EAAE,CAAC;QACb,IAAI,GAAG,YAAY,aAAa;YAAE,MAAM,GAAG,CAAC;QAC5C,MAAM,SAAS,CAAC,cAAc,EAAE,wBAAwB,QAAQ,EAAE,EAAE,GAAG,CAAC,CAAC;IAC3E,CAAC;AACH,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,4BAA4B,CAChD,GAAW;IAEX,IAAI,CAAC;QACH,MAAM,GAAG,GAAG,MAAM,KAAK,CAAC,SAAS,CAAC,GAAG,CAAC,CAAC;QACvC,OAAO,eAAe,CAAC,GAAG,CAAC,CAAC;IAC9B,CAAC;IAAC,OAAO,GAAG,EAAE,CAAC;QACb,IAAI,GAAG,YAAY,aAAa;YAAE,MAAM,GAAG,CAAC;QAC5C,MAAM,SAAS,CAAC,cAAc,EAAE,6BAA6B,EAAE,GAAG,CAAC,CAAC;IACtE,CAAC;AACH,CAAC;AAED,MAAM,UAAU,gBAAgB,CAAC,UAA0B;IACzD,IAAI,UAAU,CAAC,MAAM,EAAE,MAAM,EAAE,CAAC;QAC9B,OAAO,yBAAyB,CAAC,UAAU,CAAC,MAAM,CAAC,CAAC;IACtD,CAAC;IACD,MAAM,GAAG,GAAa,EAAE,CAAC;IACzB,UAAU,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,KAAK,EAAE,CAAC,EAAE,EAAE;QACrC,IAAI,KAAK,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;YACvB,GAAG,CAAC,IAAI,CAAC,SAAS,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;YAC3B,OAAO;QACT,CAAC;QACD,KAAK,MAAM,CAAC,IAAI,KAAK;YAAE,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;IACrC,CAAC,CAAC,CAAC;IACH,OAAO,GAAG,CAAC;AACb,CAAC;AAED,MAAM,UAAU,UAAU,CAAC,UAA0B;IACnD,OAAO,UAAU,CAAC,MAAM;SACrB,GAAG,CAAC,CAAC,KAAK,EAAE,CAAC,EAAE,EAAE;QAChB,MAAM,IAAI,GAAG,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;QAC9B,OAAO,IAAI,CAAC,CAAC,CAAC,YAAY,CAAC,GAAG,CAAC,KAAK,IAAI,EAAE,CAAC,CAAC,CAAC,YAAY,CAAC,GAAG,CAAC,EAAE,CAAC;IACnE,CAAC,CAAC;SACD,IAAI,CAAC,MAAM,CAAC;SACZ,IAAI,EAAE,CAAC;AACZ,CAAC"}
|
|
@@ -1 +1 @@
|
|
|
1
|
-
{"version":3,"file":"router.d.ts","sourceRoot":"","sources":["../../src/parse/router.ts"],"names":[],"mappings":"AAEA,OAAO,KAAK,EACV,aAAa,EACb,cAAc,EACd,YAAY,EACb,MAAM,aAAa,CAAC;
|
|
1
|
+
{"version":3,"file":"router.d.ts","sourceRoot":"","sources":["../../src/parse/router.ts"],"names":[],"mappings":"AAEA,OAAO,KAAK,EACV,aAAa,EACb,cAAc,EACd,YAAY,EACb,MAAM,aAAa,CAAC;AAuErB;;GAEG;AACH,wBAAsB,aAAa,CACjC,KAAK,EAAE,aAAa,EACpB,OAAO,GAAE,YAAiB,GACzB,OAAO,CAAC,cAAc,CAAC,CAgMzB;AA4FD,OAAO,EAAE,gBAAgB,EAAE,MAAM,iBAAiB,CAAC"}
|
package/dist/parse/router.js
CHANGED
|
@@ -7,6 +7,8 @@ import { assertSupportedParse, detectFormatFromBuffer, detectFormatFromText, for
|
|
|
7
7
|
import { extractPdfText } from './pdf.js';
|
|
8
8
|
import { pptxToParagraphs, pptxToText, readPptxStructured, readPptxStructuredFromBuffer, } from './pptx.js';
|
|
9
9
|
import { parseCsv, parseHtml, parseMarkdown, readTextFile, } from './text-formats.js';
|
|
10
|
+
import { htmlToStructured, markdownToBlocks } from './html-blocks.js';
|
|
11
|
+
import { flattenBlocksToParagraphs } from '../convert/blocks.js';
|
|
10
12
|
import { chunkText } from '../convert/emit.js';
|
|
11
13
|
function toBuffer(input) {
|
|
12
14
|
if (typeof input === 'string')
|
|
@@ -155,8 +157,17 @@ export async function parseDocument(input, options = {}) {
|
|
|
155
157
|
: '';
|
|
156
158
|
if (format === 'md') {
|
|
157
159
|
const parsed = parseMarkdown(raw);
|
|
160
|
+
const blocks = markdownToBlocks(parsed.text);
|
|
161
|
+
const paragraphs = flattenBlocksToParagraphs(blocks).length > 0
|
|
162
|
+
? flattenBlocksToParagraphs(blocks)
|
|
163
|
+
: parsed.paragraphs;
|
|
158
164
|
if (output === 'structured') {
|
|
159
|
-
const structured = {
|
|
165
|
+
const structured = {
|
|
166
|
+
format: 'md',
|
|
167
|
+
paragraphs,
|
|
168
|
+
blocks,
|
|
169
|
+
raw: parsed.text,
|
|
170
|
+
};
|
|
160
171
|
const payload = JSON.stringify(structured, null, 2);
|
|
161
172
|
return {
|
|
162
173
|
format: 'md',
|
|
@@ -167,13 +178,20 @@ export async function parseDocument(input, options = {}) {
|
|
|
167
178
|
};
|
|
168
179
|
}
|
|
169
180
|
return finish('md', parsed.text, filePath, output, {
|
|
170
|
-
paragraphs
|
|
181
|
+
paragraphs,
|
|
182
|
+
blocks,
|
|
183
|
+
raw: parsed.text,
|
|
171
184
|
});
|
|
172
185
|
}
|
|
173
186
|
if (format === 'html') {
|
|
174
|
-
const
|
|
187
|
+
const rich = htmlToStructured(raw);
|
|
188
|
+
const plain = parseHtml(raw);
|
|
175
189
|
if (output === 'structured') {
|
|
176
|
-
const structured = {
|
|
190
|
+
const structured = {
|
|
191
|
+
format: 'html',
|
|
192
|
+
paragraphs: rich.paragraphs.length ? rich.paragraphs : plain.paragraphs,
|
|
193
|
+
blocks: rich.blocks,
|
|
194
|
+
};
|
|
177
195
|
const payload = JSON.stringify(structured, null, 2);
|
|
178
196
|
return {
|
|
179
197
|
format: 'html',
|
|
@@ -183,8 +201,9 @@ export async function parseDocument(input, options = {}) {
|
|
|
183
201
|
structured,
|
|
184
202
|
};
|
|
185
203
|
}
|
|
186
|
-
return finish('html',
|
|
187
|
-
paragraphs:
|
|
204
|
+
return finish('html', plain.text, filePath, output, {
|
|
205
|
+
paragraphs: rich.paragraphs.length ? rich.paragraphs : plain.paragraphs,
|
|
206
|
+
blocks: rich.blocks,
|
|
188
207
|
});
|
|
189
208
|
}
|
|
190
209
|
const parsed = parseCsv(raw);
|