@parseo/shared 1.0.0 → 1.0.1

package/dist/cli.d.ts

@@ -0,0 +1,15 @@
+/**
+ * Resolve a file path from CLI arguments, handling drag-and-drop artifacts.
+ *
+ * When a file is dragged into a terminal on macOS, the shell may:
+ *  - Backslash-escape spaces, ampersands, parentheses, etc.
+ *  - Split the path into multiple argv entries if quoting is incomplete.
+ *  - Eat special characters like `&` even inside quotes when lines wrap.
+ *
+ * This function joins all non-flag arguments, strips leftover escape
+ * backslashes, trims surrounding quotes/whitespace, and if the exact path
+ * doesn't exist, walks up to find the deepest valid ancestor and fuzzy-matches
+ * each missing segment against the actual directory contents.
+ */
+export declare function resolvePathFromArgs(args: string[]): string;
+//# sourceMappingURL=cli.d.ts.map

package/dist/cli.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"cli.d.ts","sourceRoot":"","sources":["../src/cli.ts"],"names":[],"mappings":"AAQA;;;;;;;;;;;;GAYG;AACH,wBAAgB,mBAAmB,CAAC,IAAI,EAAE,MAAM,EAAE,GAAG,MAAM,CAuC1D"}

package/dist/cli.js

@@ -0,0 +1,86 @@
+import { existsSync, readdirSync } from "fs";
+import { resolve, join } from "path";
+/**
+ * Flags that consume the next argument as their value (not part of the path).
+ */
+const FLAGS_WITH_VALUES = new Set(["--output"]);
+/**
+ * Resolve a file path from CLI arguments, handling drag-and-drop artifacts.
+ *
+ * When a file is dragged into a terminal on macOS, the shell may:
+ *  - Backslash-escape spaces, ampersands, parentheses, etc.
+ *  - Split the path into multiple argv entries if quoting is incomplete.
+ *  - Eat special characters like `&` even inside quotes when lines wrap.
+ *
+ * This function joins all non-flag arguments, strips leftover escape
+ * backslashes, trims surrounding quotes/whitespace, and if the exact path
+ * doesn't exist, walks up to find the deepest valid ancestor and fuzzy-matches
+ * each missing segment against the actual directory contents.
+ */
+export function resolvePathFromArgs(args) {
+    const pathParts = [];
+    for (let i = 0; i < args.length; i++) {
+        if (FLAGS_WITH_VALUES.has(args[i])) {
+            i++; // skip the flag's value
+        }
+        else if (!args[i].startsWith("--")) {
+            pathParts.push(args[i]);
+        }
+    }
+    if (pathParts.length === 0) {
+        return "";
+    }
+    let raw = pathParts.join(" ");
+    // Strip surrounding single or double quotes
+    raw = raw.replace(/^(['"])(.*)\1$/, "$2");
+    // Remove backslash escapes added by drag-and-drop (e.g. `\ `, `\&`, `\(`)
+    raw = raw.replace(/\\(?=[ &()'])/g, "");
+    raw = raw.trim();
+    const resolved = resolve(raw);
+    if (existsSync(resolved)) {
+        return resolved;
+    }
+    // Fuzzy resolve: walk the path segments and match against disk contents
+    // when a segment doesn't exist (e.g. shell ate `&` from a folder name).
+    const fuzzyResolved = fuzzyResolvePath(resolved);
+    if (fuzzyResolved) {
+        return fuzzyResolved;
+    }
+    console.error(`File not found: ${resolved}`);
+    process.exit(1);
+}
+/**
+ * Walk the path from root to leaf. At each level, if the segment doesn't
+ * match an entry exactly, find the entry whose name matches after collapsing
+ * whitespace and stripping punctuation (handles missing `&`, extra spaces, etc).
+ */
+function fuzzyResolvePath(fullPath) {
+    const segments = fullPath.split("/");
+    let current = "/";
+    for (let i = 1; i < segments.length; i++) {
+        const seg = segments[i];
+        if (!seg)
+            continue;
+        const exact = join(current, seg);
+        if (existsSync(exact)) {
+            current = exact;
+            continue;
+        }
+        // Segment doesn't exist — try fuzzy match against directory entries
+        let entries;
+        try {
+            entries = readdirSync(current);
+        }
+        catch {
+            return null;
+        }
+        const normalize = (s) => s.toLowerCase().replace(/[^a-z0-9]/g, "");
+        const target = normalize(seg);
+        const match = entries.find((e) => normalize(e) === target);
+        if (!match) {
+            return null;
+        }
+        current = join(current, match);
+    }
+    return existsSync(current) ? current : null;
+}

package/dist/index.d.ts

@@ -1,3 +1,4 @@
+export { resolvePathFromArgs } from "./cli.js";
 export { extractTextItems, extractLines, formLines, extractFilledRects } from "./extract.js";
 export type { FilledRect } from "./extract.js";
 export { classifyDocument } from "./classify.js";
@@ -5,7 +6,6 @@ export type { FormatName, ClassifyResult, PackageName } from "./classify.js";
 export type { TextItem, TextSegment, TextLine, DateString, DateRange } from "./types.js";
 export { toBBox } from "./types.js";
 export type { BoundingBox } from "./types.js";
-export { resolvePathFromArgs } from "./cli.js";
 export { ParserError, InvalidPDFError, UnrecognizedFormatError, MissingSectionError, ExtractionError, } from "./errors.js";
 export { parseDate, parseDateRange, parseCurrency, parseNum, escapeRegex, cleanNumber, getSegmentNear, extractLabelValue, findLabelInText, isBulletLine, isNumberedEntry, parseBulletKeyValues, collectBulletItems, collectUntil, mapToColumns, findColumnHeaders, getSection, } from "./utils.js";
 export type { Section } from "./utils.js";

package/dist/index.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,gBAAgB,EAAE,YAAY,EAAE,SAAS,EAAE,kBAAkB,EAAE,MAAM,cAAc,CAAC;AAC7F,YAAY,EAAE,UAAU,EAAE,MAAM,cAAc,CAAC;AAC/C,OAAO,EAAE,gBAAgB,EAAE,MAAM,eAAe,CAAC;AACjD,YAAY,EAAE,UAAU,EAAE,cAAc,EAAE,WAAW,EAAE,MAAM,eAAe,CAAC;AAC7E,YAAY,EAAE,QAAQ,EAAE,WAAW,EAAE,QAAQ,EAAE,UAAU,EAAE,SAAS,EAAE,MAAM,YAAY,CAAC;AACzF,OAAO,EAAE,MAAM,EAAE,MAAM,YAAY,CAAC;AACpC,YAAY,EAAE,WAAW,EAAE,MAAM,YAAY,CAAC;AAC9C,OAAO,EAAE,mBAAmB,EAAE,MAAM,UAAU,CAAC;AAC/C,OAAO,EACL,WAAW,EACX,eAAe,EACf,uBAAuB,EACvB,mBAAmB,EACnB,eAAe,GAChB,MAAM,aAAa,CAAC;AACrB,OAAO,EACL,SAAS,EACT,cAAc,EACd,aAAa,EACb,QAAQ,EACR,WAAW,EACX,WAAW,EACX,cAAc,EACd,iBAAiB,EACjB,eAAe,EACf,YAAY,EACZ,eAAe,EACf,oBAAoB,EACpB,kBAAkB,EAClB,YAAY,EACZ,YAAY,EACZ,iBAAiB,EACjB,UAAU,GACX,MAAM,YAAY,CAAC;AACpB,YAAY,EAAE,OAAO,EAAE,MAAM,YAAY,CAAC"}
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,mBAAmB,EAAE,MAAM,UAAU,CAAC;AAC/C,OAAO,EAAE,gBAAgB,EAAE,YAAY,EAAE,SAAS,EAAE,kBAAkB,EAAE,MAAM,cAAc,CAAC;AAC7F,YAAY,EAAE,UAAU,EAAE,MAAM,cAAc,CAAC;AAC/C,OAAO,EAAE,gBAAgB,EAAE,MAAM,eAAe,CAAC;AACjD,YAAY,EAAE,UAAU,EAAE,cAAc,EAAE,WAAW,EAAE,MAAM,eAAe,CAAC;AAC7E,YAAY,EAAE,QAAQ,EAAE,WAAW,EAAE,QAAQ,EAAE,UAAU,EAAE,SAAS,EAAE,MAAM,YAAY,CAAC;AACzF,OAAO,EAAE,MAAM,EAAE,MAAM,YAAY,CAAC;AACpC,YAAY,EAAE,WAAW,EAAE,MAAM,YAAY,CAAC;AAC9C,OAAO,EACL,WAAW,EACX,eAAe,EACf,uBAAuB,EACvB,mBAAmB,EACnB,eAAe,GAChB,MAAM,aAAa,CAAC;AACrB,OAAO,EACL,SAAS,EACT,cAAc,EACd,aAAa,EACb,QAAQ,EACR,WAAW,EACX,WAAW,EACX,cAAc,EACd,iBAAiB,EACjB,eAAe,EACf,YAAY,EACZ,eAAe,EACf,oBAAoB,EACpB,kBAAkB,EAClB,YAAY,EACZ,YAAY,EACZ,iBAAiB,EACjB,UAAU,GACX,MAAM,YAAY,CAAC;AACpB,YAAY,EAAE,OAAO,EAAE,MAAM,YAAY,CAAC"}

package/dist/index.js

@@ -1,6 +1,6 @@
+export { resolvePathFromArgs } from "./cli.js";
 export { extractTextItems, extractLines, formLines, extractFilledRects } from "./extract.js";
 export { classifyDocument } from "./classify.js";
 export { toBBox } from "./types.js";
-export { resolvePathFromArgs } from "./cli.js";
 export { ParserError, InvalidPDFError, UnrecognizedFormatError, MissingSectionError, ExtractionError, } from "./errors.js";
 export { parseDate, parseDateRange, parseCurrency, parseNum, escapeRegex, cleanNumber, getSegmentNear, extractLabelValue, findLabelInText, isBulletLine, isNumberedEntry, parseBulletKeyValues, collectBulletItems, collectUntil, mapToColumns, findColumnHeaders, getSection, } from "./utils.js";

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@parseo/shared",
-  "version": "1.0.0",
+  "version": "1.0.1",
   "type": "module",
   "main": "./dist/index.js",
   "types": "./dist/index.d.ts",
@@ -14,7 +14,9 @@
   "publishConfig": {
     "access": "public"
   },
-  "files": ["dist", "!dist/**/cli.*"],
+  "files": [
+    "dist"
+  ],
   "scripts": {
     "build": "tsc"
   },