Blame view

node_modules/recast/lib/parser.js 8.47 KB
ce4c83ff   wxy   初始提交
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
  "use strict";
  
  var assert = require("assert");
  var types = require("./types");
  var n = types.namedTypes;
  var b = types.builders;
  var isObject = types.builtInTypes.object;
  var isArray = types.builtInTypes.array;
  var isFunction = types.builtInTypes.function;
  var Patcher = require("./patcher").Patcher;
  var normalizeOptions = require("./options").normalize;
  var fromString = require("./lines").fromString;
  var attachComments = require("./comments").attach;
  var util = require("./util");
  
  exports.parse = function parse(source, options) {
    options = normalizeOptions(options);
  
    const lines = fromString(source, options);
  
    const sourceWithoutTabs = lines.toString({
      tabWidth: options.tabWidth,
      reuseWhitespace: false,
      useTabs: false
    });
  
    let comments = [];
    const ast = options.parser.parse(sourceWithoutTabs, {
      jsx: true,
      loc: true,
      locations: true,
      range: options.range,
      comment: true,
      onComment: comments,
      tolerant: util.getOption(options, "tolerant", true),
      ecmaVersion: 6,
      sourceType: util.getOption(options, "sourceType", "module")
    });
  
    // Use ast.tokens if possible, and otherwise fall back to the Esprima
    // tokenizer. All the preconfigured ../parsers/* expose ast.tokens
    // automatically, but custom parsers might need additional configuration
    // to avoid this fallback.
    const tokens = Array.isArray(ast.tokens)
      ? ast.tokens
      : require("esprima").tokenize(sourceWithoutTabs, {
          loc: true
        });
  
    // We will reattach the tokens array to the file object below.
    delete ast.tokens;
  
    // Make sure every token has a token.value string.
    tokens.forEach(function (token) {
      if (typeof token.value !== "string") {
        token.value = lines.sliceString(token.loc.start, token.loc.end);
      }
    });
  
    if (Array.isArray(ast.comments)) {
      comments = ast.comments;
      delete ast.comments;
    }
  
    if (ast.loc) {
      // If the source was empty, some parsers give loc.{start,end}.line
      // values of 0, instead of the minimum of 1.
      util.fixFaultyLocations(ast, lines);
    } else {
      ast.loc = {
        start: lines.firstPos(),
        end: lines.lastPos()
      };
    }
  
    ast.loc.lines = lines;
    ast.loc.indent = 0;
  
    let file;
    let program;
    if (ast.type === "Program") {
      program = ast;
      // In order to ensure we reprint leading and trailing program
      // comments, wrap the original Program node with a File node. Only
      // ESTree parsers (Acorn and Esprima) return a Program as the root AST
      // node. Most other (Babylon-like) parsers return a File.
      file = b.file(ast, options.sourceFileName || null);
      file.loc = {
        start: lines.firstPos(),
        end: lines.lastPos(),
        lines: lines,
        indent: 0
      };
    } else if (ast.type === "File") {
      file = ast;
      program = file.program;
    }
  
    // Expose file.tokens unless the caller passed false for options.tokens.
    if (options.tokens) {
      file.tokens = tokens;
    }
  
    // Expand the Program's .loc to include all comments (not just those
    // attached to the Program node, as its children may have comments as
    // well), since sometimes program.loc.{start,end} will coincide with the
    // .loc.{start,end} of the first and last *statements*, mistakenly
    // excluding comments that fall outside that region.
    var trueProgramLoc = util.getTrueLoc({
      type: program.type,
      loc: program.loc,
      body: [],
      comments
    }, lines);
    program.loc.start = trueProgramLoc.start;
    program.loc.end = trueProgramLoc.end;
  
    // Passing file.program here instead of just file means that initial
    // comments will be attached to program.body[0] instead of program.
    attachComments(
      comments,
      program.body.length ? file.program : file,
      lines
    );
  
    // Return a copy of the original AST so that any changes made may be
    // compared to the original.
    return new TreeCopier(lines, tokens).copy(file);
  };
  
  function TreeCopier(lines, tokens) {
    assert.ok(this instanceof TreeCopier);
    this.lines = lines;
    this.tokens = tokens;
    this.startTokenIndex = 0;
    this.endTokenIndex = tokens.length;
    this.indent = 0;
    this.seen = new Map;
  }
  
  var TCp = TreeCopier.prototype;
  
  TCp.copy = function(node) {
    if (this.seen.has(node)) {
      return this.seen.get(node);
    }
  
    if (isArray.check(node)) {
      var copy = new Array(node.length);
      this.seen.set(node, copy);
      node.forEach(function (item, i) {
        copy[i] = this.copy(item);
      }, this);
      return copy;
    }
  
    if (!isObject.check(node)) {
      return node;
    }
  
    util.fixFaultyLocations(node, this.lines);
  
    var copy = Object.create(Object.getPrototypeOf(node), {
      original: { // Provide a link from the copy to the original.
        value: node,
        configurable: false,
        enumerable: false,
        writable: true
      }
    });
  
    this.seen.set(node, copy);
  
    var loc = node.loc;
    var oldIndent = this.indent;
    var newIndent = oldIndent;
  
    const oldStartTokenIndex = this.startTokenIndex;
    const oldEndTokenIndex = this.endTokenIndex;
  
    if (loc) {
      // When node is a comment, we set node.loc.indent to
      // node.loc.start.column so that, when/if we print the comment by
      // itself, we can strip that much whitespace from the left margin of
      // the comment. This only really matters for multiline Block comments,
      // but it doesn't hurt for Line comments.
      if (node.type === "Block" || node.type === "Line" ||
          node.type === "CommentBlock" || node.type === "CommentLine" ||
          this.lines.isPrecededOnlyByWhitespace(loc.start)) {
        newIndent = this.indent = loc.start.column;
      }
  
      // Every node.loc has a reference to the original source lines as well
      // as a complete list of source tokens.
      loc.lines = this.lines;
      loc.tokens = this.tokens;
      loc.indent = newIndent;
  
      // Set loc.start.token and loc.end.token such that
      // loc.tokens.slice(loc.start.token, loc.end.token) returns a list of
      // all the tokens that make up this node.
      this.findTokenRange(loc);
    }
  
    var keys = Object.keys(node);
    var keyCount = keys.length;
    for (var i = 0; i < keyCount; ++i) {
      var key = keys[i];
      if (key === "loc") {
        copy[key] = node[key];
      } else if (key === "tokens" &&
                 node.type === "File") {
        // Preserve file.tokens (uncopied) in case client code cares about
        // it, even though Recast ignores it when reprinting.
        copy[key] = node[key];
      } else {
        copy[key] = this.copy(node[key]);
      }
    }
  
    this.indent = oldIndent;
    this.startTokenIndex = oldStartTokenIndex;
    this.endTokenIndex = oldEndTokenIndex;
  
    return copy;
  };
  
  // If we didn't have any idea where in loc.tokens to look for tokens
  // contained by this loc, a binary search would be appropriate, but
  // because we maintain this.startTokenIndex and this.endTokenIndex as we
  // traverse the AST, we only need to make small (linear) adjustments to
  // those indexes with each recursive iteration.
  TCp.findTokenRange = function (loc) {
    // In the unlikely event that loc.tokens[this.startTokenIndex] starts
    // *after* loc.start, we need to rewind this.startTokenIndex first.
    while (this.startTokenIndex > 0) {
      const token = loc.tokens[this.startTokenIndex];
      if (util.comparePos(loc.start, token.loc.start) < 0) {
        --this.startTokenIndex;
      } else break;
    }
  
    // In the unlikely event that loc.tokens[this.endTokenIndex - 1] ends
    // *before* loc.end, we need to fast-forward this.endTokenIndex first.
    while (this.endTokenIndex < loc.tokens.length) {
      const token = loc.tokens[this.endTokenIndex];
      if (util.comparePos(token.loc.end, loc.end) < 0) {
        ++this.endTokenIndex;
      } else break;
    }
  
    // Increment this.startTokenIndex until we've found the first token
    // contained by this node.
    while (this.startTokenIndex < this.endTokenIndex) {
      const token = loc.tokens[this.startTokenIndex];
      if (util.comparePos(token.loc.start, loc.start) < 0) {
        ++this.startTokenIndex;
      } else break;
    }
  
    // Index into loc.tokens of the first token within this node.
    loc.start.token = this.startTokenIndex;
  
    // Decrement this.endTokenIndex until we've found the first token after
    // this node (not contained by the node).
    while (this.endTokenIndex > this.startTokenIndex) {
      const token = loc.tokens[this.endTokenIndex - 1];
      if (util.comparePos(loc.end, token.loc.end) < 0) {
        --this.endTokenIndex;
      } else break;
    }
  
    // Index into loc.tokens of the first token *after* this node.
    // If loc.start.token === loc.end.token, the node contains no tokens,
    // and the index is that of the next token following this node.
    loc.end.token = this.endTokenIndex;
  };