node-opcua-xml2json 2.176.0 → 2.181.0

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.
Files changed (41) hide show
  1. package/dist/source/definition_parser.d.ts +1 -1
  2. package/dist/source/definition_parser.js +3 -3
  3. package/dist/source/definition_parser.js.map +1 -1
  4. package/dist/source/fragment_cloner.d.ts +5 -5
  5. package/dist/source/fragment_cloner.js +11 -8
  6. package/dist/source/fragment_cloner.js.map +1 -1
  7. package/dist/source/fragment_cloner_parser.d.ts +2 -2
  8. package/dist/source/fragment_cloner_parser.js +7 -5
  9. package/dist/source/fragment_cloner_parser.js.map +1 -1
  10. package/dist/source/index.d.ts +5 -5
  11. package/dist/source/index.js +5 -5
  12. package/dist/source/index.js.map +1 -1
  13. package/dist/source/nodejs/xml2json_fs.d.ts +2 -2
  14. package/dist/source/nodejs/xml2json_fs.js +4 -4
  15. package/dist/source/nodejs/xml2json_fs.js.map +1 -1
  16. package/dist/source/thirdparties/escape.js +9 -14
  17. package/dist/source/thirdparties/escape.js.map +1 -1
  18. package/dist/source/thirdparties/parser/lts.d.ts +18 -2
  19. package/dist/source/thirdparties/parser/lts.js +110 -21
  20. package/dist/source/thirdparties/parser/lts.js.map +1 -1
  21. package/dist/source/xml2Json_pojo_tools.d.ts +1 -1
  22. package/dist/source/xml2Json_pojo_tools.js +2 -2
  23. package/dist/source/xml2Json_pojo_tools.js.map +1 -1
  24. package/dist/source/xml2json.d.ts +22 -38
  25. package/dist/source/xml2json.js +189 -86
  26. package/dist/source/xml2json.js.map +1 -1
  27. package/dist/source/xml2json_pojo.d.ts +2 -2
  28. package/dist/source/xml2json_pojo.js +9 -6
  29. package/dist/source/xml2json_pojo.js.map +1 -1
  30. package/dist/tsconfig.tsbuildinfo +1 -1
  31. package/package.json +8 -7
  32. package/source/definition_parser.ts +9 -5
  33. package/source/fragment_cloner.ts +8 -7
  34. package/source/fragment_cloner_parser.ts +8 -6
  35. package/source/index.ts +5 -5
  36. package/source/nodejs/xml2json_fs.ts +3 -3
  37. package/source/thirdparties/escape.ts +9 -20
  38. package/source/thirdparties/parser/lts.ts +133 -38
  39. package/source/xml2Json_pojo_tools.ts +3 -5
  40. package/source/xml2json.ts +219 -98
  41. package/source/xml2json_pojo.ts +8 -5
@@ -1,7 +1,6 @@
1
- 'use strict';
2
-
3
- import { EventEmitter } from "events";
4
- import { unescapeXML } from "../escape";
1
+ import { EventEmitter } from "node:events";
2
+ import type { XmlAttributes } from "../../xml2json.js";
3
+ import { unescapeXML } from "../escape.js";
5
4
 
6
5
  const STATE_TEXT = 0;
7
6
  const STATE_IGNORE_COMMENT = 1;
@@ -15,41 +14,96 @@ const STATE_ATTR_VALUE = 8;
15
14
  const STATE_CDATA = 9;
16
15
  const STATE_IGNORE_CDATA = 10;
17
16
 
18
- export class SaxLtx extends EventEmitter {
17
+ // where a tag name or an attribute name ends: found by the regex engine in one step rather than by
18
+ // the loop below one character at a time (the way text runs and attribute values already are)
19
+ const TAG_NAME_END = /[\s/>]/g;
20
+ const ATTR_NAME_END = /[\s=]/g;
21
+
22
+ /**
23
+ * Handlers called directly, without going through EventEmitter. A nodeset carries hundreds of
24
+ * thousands of elements and text runs, and the emit machinery (listener lookup, arguments array,
25
+ * apply) was a measurable share of the parse; the events are still emitted when no handler is set.
26
+ */
27
+ export interface SaxLtxHandlers {
28
+ onStartElement?: (name: string, attrs: XmlAttributes) => void;
29
+ onEndElement?: (name: string, selfClosing: boolean) => void;
30
+ onText?: (text: string) => void;
31
+ /**
32
+ * drop a text run made of white space only before unescaping it: the indentation between
33
+ * two tags, which is most of the text runs of a pretty-printed document
34
+ */
35
+ skipBlankText?: boolean;
36
+ }
19
37
 
38
+ function isBlank(text: string): boolean {
39
+ for (let i = 0; i < text.length; i++) {
40
+ if (text.charCodeAt(i) > 32) {
41
+ return false;
42
+ }
43
+ }
44
+ return true;
45
+ }
20
46
 
47
+ export class SaxLtx extends EventEmitter {
21
48
  #write: (data: string) => void;
22
- constructor() {
49
+ constructor(handlers: SaxLtxHandlers = {}) {
23
50
  super();
24
51
 
52
+ const onStartElement = handlers.onStartElement;
53
+ const onEndElement = handlers.onEndElement;
54
+ const onText = handlers.onText;
55
+ const skipBlankText = !!handlers.skipBlankText;
56
+
57
+ const emitText = (text: string) => {
58
+ if (skipBlankText && isBlank(text)) {
59
+ return;
60
+ }
61
+ const unescaped = unescapeXML(text);
62
+ if (onText) {
63
+ onText(unescaped);
64
+ } else {
65
+ this.emit("text", unescaped);
66
+ }
67
+ };
68
+
25
69
  function _handleTagOpening(
26
70
  this: SaxLtx,
27
- endTag: boolean| undefined,
28
- tagName: string|undefined,
29
- attrs: string
71
+ endTag: boolean | undefined,
72
+ tagName: string | undefined,
73
+ attrs: XmlAttributes | undefined
30
74
  ) {
31
75
  if (!endTag) {
32
- this.emit("startElement", tagName, attrs);
76
+ if (onStartElement) {
77
+ onStartElement(tagName as string, attrs as XmlAttributes);
78
+ } else {
79
+ this.emit("startElement", tagName, attrs);
80
+ }
33
81
  if (selfClosing) {
34
- this.emit("endElement", tagName, true);
82
+ if (onEndElement) {
83
+ onEndElement(tagName as string, true);
84
+ } else {
85
+ this.emit("endElement", tagName, true);
86
+ }
35
87
  }
88
+ } else if (onEndElement) {
89
+ onEndElement(tagName as string, false);
36
90
  } else {
37
91
  this.emit("endElement", tagName, false);
38
92
  }
39
- };
93
+ }
40
94
  let state = STATE_TEXT;
41
95
  let remainder: string | null = null;
42
96
  let parseRemainder: boolean = false;
43
97
  let tagName: string | undefined;
44
- let attrs: string | {} | undefined;
45
- let endTag: boolean|undefined;
98
+ let attrs: XmlAttributes | undefined;
99
+ let endTag: boolean | undefined;
46
100
  let selfClosing: boolean | undefined;
47
101
  let attrQuote: number;
48
102
  let attrQuoteChar: string;
49
103
  let recordStart: number | undefined = 0;
50
- let attrName: string|undefined;
104
+ let attrName: string | undefined;
51
105
 
52
- this.#write = function write(data: string ) {
106
+ this.#write = function write(data: string) {
53
107
  let pos = 0;
54
108
 
55
109
  /* Anything from previous write()? */
@@ -66,29 +120,59 @@ export class SaxLtx extends EventEmitter {
66
120
  recordStart = undefined;
67
121
  return recorded;
68
122
  }
69
- return undefined;
123
+ return undefined;
70
124
  }
71
125
 
126
+ // Each fast-forward below looks for the end of the record being read. When the chunk
127
+ // holds no such end, the rest of the chunk belongs to the record: skip to the end at
128
+ // once. Searching again from every following character made a chunk cost the square
129
+ // of its unread tail, 6 seconds for a 4 MB text run delivered in 64 KB pieces.
72
130
  for (; pos < data.length; pos++) {
73
131
  switch (state) {
74
132
  case STATE_TEXT: {
75
133
  // if we're looping through text, fast-forward using indexOf to
76
134
  // the next '<' character
77
135
  const lt = data.indexOf("<", pos);
78
- if (lt !== -1 && pos !== lt) {
79
- pos = lt;
136
+ if (lt === -1) {
137
+ pos = data.length;
138
+ continue;
80
139
  }
81
-
140
+ pos = lt;
82
141
  break;
83
142
  }
84
143
  case STATE_ATTR_VALUE: {
85
144
  // if we're looping through an attribute, fast-forward using
86
145
  // indexOf to the next end quote character
87
146
  const quot = data.indexOf(attrQuoteChar, pos);
88
- if (quot !== -1) {
89
- pos = quot;
147
+ if (quot === -1) {
148
+ pos = data.length;
149
+ continue;
90
150
  }
91
-
151
+ pos = quot;
152
+ break;
153
+ }
154
+ case STATE_TAG_NAME: {
155
+ // past the first character, which may be "/", "!" or "?", fast-forward
156
+ // to the end of the name
157
+ if (recordStart !== pos) {
158
+ TAG_NAME_END.lastIndex = pos;
159
+ const m = TAG_NAME_END.exec(data);
160
+ if (!m) {
161
+ pos = data.length;
162
+ continue;
163
+ }
164
+ pos = m.index;
165
+ }
166
+ break;
167
+ }
168
+ case STATE_ATTR_NAME: {
169
+ ATTR_NAME_END.lastIndex = pos;
170
+ const m = ATTR_NAME_END.exec(data);
171
+ if (!m) {
172
+ pos = data.length;
173
+ continue;
174
+ }
175
+ pos = m.index;
92
176
  break;
93
177
  }
94
178
  case STATE_IGNORE_COMMENT: {
@@ -120,7 +204,7 @@ export class SaxLtx extends EventEmitter {
120
204
  if (c === 60 /* < */) {
121
205
  const text = endRecording();
122
206
  if (text) {
123
- this.emit("text", unescapeXML(text));
207
+ emitText(text);
124
208
  }
125
209
  state = STATE_TAG_NAME;
126
210
  recordStart = pos + 1;
@@ -129,13 +213,21 @@ export class SaxLtx extends EventEmitter {
129
213
  break;
130
214
  case STATE_CDATA:
131
215
  if (c === 93 /* ] */) {
132
- if (data.substring(pos + 1, 2) === "]>") {
216
+ if (data.startsWith("]>", pos + 1)) {
133
217
  const cData = endRecording();
134
218
  if (cData) {
135
- this.emit("text", cData);
219
+ if (onText) {
220
+ onText(cData);
221
+ } else {
222
+ this.emit("text", cData);
223
+ }
136
224
  }
225
+ // skip the "]>" that closes the section
226
+ pos += 2;
137
227
  state = STATE_TEXT;
138
- } else if (data.length < pos + 2) {
228
+ recordStart = pos + 1;
229
+ } else if (data.length < pos + 3) {
230
+ // the closing "]]>" may straddle this chunk and the next
139
231
  parseRemainder = true;
140
232
  pos = data.length;
141
233
  }
@@ -146,13 +238,10 @@ export class SaxLtx extends EventEmitter {
146
238
  recordStart = pos + 1;
147
239
  endTag = true;
148
240
  } else if (c === 33 /* ! */) {
149
- if (data.substring(pos + 1, 7) === "[CDATA[") {
241
+ if (data.startsWith("[CDATA[", pos + 1)) {
150
242
  recordStart = pos + 8;
151
243
  state = STATE_CDATA;
152
- } else if (
153
- data.length < pos + 8 &&
154
- "[CDATA[".startsWith(data.slice(pos + 1))
155
- ) {
244
+ } else if (data.length < pos + 8 && "[CDATA[".startsWith(data.slice(pos + 1))) {
156
245
  // We potentially have CDATA, but the chunk is ending; stop here and let the next write() decide
157
246
  parseRemainder = true;
158
247
  pos = data.length;
@@ -191,7 +280,7 @@ export class SaxLtx extends EventEmitter {
191
280
  break;
192
281
  case STATE_TAG:
193
282
  if (c === 62 /* > */) {
194
- _handleTagOpening.call(this, endTag, tagName, attrs as any);
283
+ _handleTagOpening.call(this, endTag, tagName, attrs);
195
284
  tagName = undefined;
196
285
  attrs = undefined;
197
286
  endTag = undefined;
@@ -227,8 +316,10 @@ export class SaxLtx extends EventEmitter {
227
316
  break;
228
317
  case STATE_ATTR_VALUE:
229
318
  if (c === attrQuote) {
230
- const value = unescapeXML(endRecording()!);
231
- (attrs as any)[attrName!] = value;
319
+ const recorded = endRecording();
320
+ if (recorded !== undefined && attrName !== undefined) {
321
+ (attrs as XmlAttributes)[attrName] = unescapeXML(recorded);
322
+ }
232
323
  attrName = undefined;
233
324
  state = STATE_TAG;
234
325
  }
@@ -239,11 +330,16 @@ export class SaxLtx extends EventEmitter {
239
330
  if (typeof recordStart === "number" && recordStart <= data.length) {
240
331
  remainder = data.slice(recordStart);
241
332
  recordStart = 0;
333
+ } else if (state === STATE_IGNORE_COMMENT || state === STATE_IGNORE_INSTRUCTION || state === STATE_IGNORE_CDATA) {
334
+ // the end of a comment (-->), of an instruction (?>) or of a skipped CDATA (]]>) is
335
+ // recognised by looking back at the characters before the '>': keep the last two
336
+ // so that a terminator straddling this chunk and the next is still seen
337
+ remainder = data.slice(-2);
338
+ parseRemainder = true;
242
339
  }
243
340
  };
244
341
  }
245
342
 
246
-
247
343
  public write(data: string | Buffer) {
248
344
  this.#write(data.toString());
249
345
  }
@@ -253,7 +349,6 @@ export class SaxLtx extends EventEmitter {
253
349
  }
254
350
 
255
351
  /* Uh, yeah */
256
- this.write = function write() { };
352
+ this.write = function write() {};
257
353
  }
258
354
  }
259
-
@@ -1,17 +1,16 @@
1
1
  import { lowerFirstLetter } from "node-opcua-utils";
2
- import { type IReaderState, ReaderStateBase, type Xml2Json, type XmlAttributes } from "./xml2json";
2
+ import { type IReaderState, ReaderStateBase, type Xml2Json, type XmlAttributes } from "./xml2json.js";
3
3
  export type withPojoLambda = (name: string, pojo: unknown) => void;
4
4
 
5
5
  export class ReaderState2 extends ReaderStateBase {
6
6
  public _stack: unknown[];
7
7
  public _pojo: unknown;
8
-
8
+
9
9
  // biome-ignore lint/suspicious/noExplicitAny: explanation
10
10
  public _element: any;
11
11
  public text: string;
12
12
  public _withPojo: withPojoLambda;
13
13
 
14
-
15
14
  private initLevel = 0;
16
15
  constructor() {
17
16
  super();
@@ -25,11 +24,10 @@ export class ReaderState2 extends ReaderStateBase {
25
24
  }
26
25
 
27
26
  public _on_init(_elementName: string, _attrs: XmlAttributes, _parent: IReaderState, level: number, _engine: Xml2Json): void {
28
-
29
27
  this.initLevel = level;
30
28
  if (this._stack.length === 0) {
31
29
  this._pojo = {};
32
- this._element = this._pojo as Record<string, unknown> ;
30
+ this._element = this._pojo as Record<string, unknown>;
33
31
  }
34
32
  }
35
33