/src/xpdf-4.06/xpdf/PSTokenizer.cc
Line | Count | Source |
1 | | //======================================================================== |
2 | | // |
3 | | // PSTokenizer.cc |
4 | | // |
5 | | // Copyright 2002-2003 Glyph & Cog, LLC |
6 | | // |
7 | | //======================================================================== |
8 | | |
9 | | #include <aconf.h> |
10 | | |
11 | | #include <stdio.h> |
12 | | #include <stdlib.h> |
13 | | #include "gmempp.h" |
14 | | #include "PSTokenizer.h" |
15 | | |
16 | | //------------------------------------------------------------------------ |
17 | | |
18 | | // A '1' in this array means the character is white space. A '1' or |
19 | | // '2' means the character ends a name or command. |
20 | | static char specialChars[256] = { |
21 | | 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 1, 0, 0, // 0x |
22 | | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 1x |
23 | | 1, 0, 0, 0, 0, 2, 0, 0, 2, 2, 0, 0, 0, 0, 0, 2, // 2x |
24 | | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 2, 0, // 3x |
25 | | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 4x |
26 | | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 2, 0, 0, // 5x |
27 | | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 6x |
28 | | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 2, 0, 0, // 7x |
29 | | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 8x |
30 | | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // 9x |
31 | | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // ax |
32 | | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // bx |
33 | | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // cx |
34 | | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // dx |
35 | | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, // ex |
36 | | 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 // fx |
37 | | }; |
38 | | |
39 | | //------------------------------------------------------------------------ |
40 | | |
41 | 21.1k | PSTokenizer::PSTokenizer(int (*getCharFuncA)(void *), void *dataA) { |
42 | 21.1k | getCharFunc = getCharFuncA; |
43 | 21.1k | data = dataA; |
44 | 21.1k | charBuf = -1; |
45 | 21.1k | } |
46 | | |
47 | 21.1k | PSTokenizer::~PSTokenizer() { |
48 | 21.1k | } |
49 | | |
50 | 1.37M | GBool PSTokenizer::getToken(char *buf, int size, int *length) { |
51 | 1.37M | GBool comment, backslash; |
52 | 1.37M | int c; |
53 | 1.37M | int i; |
54 | | |
55 | | // skip whitespace and comments |
56 | 1.37M | comment = gFalse; |
57 | 21.2M | while (1) { |
58 | 21.2M | if ((c = getChar()) == EOF) { |
59 | 24.0k | buf[0] = '\0'; |
60 | 24.0k | *length = 0; |
61 | 24.0k | return gFalse; |
62 | 24.0k | } |
63 | 21.2M | if (comment) { |
64 | 12.4M | if (c == '\x0a' || c == '\x0d') { |
65 | 16.5k | comment = gFalse; |
66 | 16.5k | } |
67 | 12.4M | } else if (c == '%') { |
68 | 16.5k | comment = gTrue; |
69 | 8.75M | } else if (specialChars[c] != 1) { |
70 | 1.35M | break; |
71 | 1.35M | } |
72 | 21.2M | } |
73 | | |
74 | | // read a token |
75 | 1.35M | i = 0; |
76 | 1.35M | buf[i++] = (char)c; |
77 | 1.35M | if (c == '(') { |
78 | 2.60k | backslash = gFalse; |
79 | 774k | while ((c = lookChar()) != EOF) { |
80 | 772k | if (i < size - 1) { |
81 | 436k | buf[i++] = (char)c; |
82 | 436k | } |
83 | 772k | getChar(); |
84 | 772k | if (c == '\\') { |
85 | 438 | backslash = gTrue; |
86 | 772k | } else if (!backslash && c == ')') { |
87 | 1.17k | break; |
88 | 771k | } else { |
89 | 771k | backslash = gFalse; |
90 | 771k | } |
91 | 772k | } |
92 | 1.34M | } else if (c == '<') { |
93 | 4.32M | while ((c = lookChar()) != EOF) { |
94 | 4.30M | getChar(); |
95 | 4.30M | if (i < size - 1 && specialChars[c] != 1) { |
96 | 3.34M | buf[i++] = (char)c; |
97 | 3.34M | } |
98 | 4.30M | if (c == '>') { |
99 | 166k | break; |
100 | 166k | } |
101 | 4.30M | } |
102 | 1.16M | } else if (c != '[' && c != ']') { |
103 | 5.33M | while ((c = lookChar()) != EOF && !specialChars[c]) { |
104 | 4.21M | getChar(); |
105 | 4.21M | if (i < size - 1) { |
106 | 3.52M | buf[i++] = (char)c; |
107 | 3.52M | } |
108 | 4.21M | } |
109 | 1.11M | } |
110 | 1.35M | buf[i] = '\0'; |
111 | 1.35M | *length = i; |
112 | | |
113 | 1.35M | return gTrue; |
114 | 1.37M | } |
115 | | |
116 | 10.4M | int PSTokenizer::lookChar() { |
117 | 10.4M | if (charBuf < 0) { |
118 | 10.4M | charBuf = (*getCharFunc)(data); |
119 | 10.4M | } |
120 | 10.4M | return charBuf; |
121 | 10.4M | } |
122 | | |
123 | 30.5M | int PSTokenizer::getChar() { |
124 | 30.5M | int c; |
125 | | |
126 | 30.5M | if (charBuf < 0) { |
127 | 20.1M | charBuf = (*getCharFunc)(data); |
128 | 20.1M | } |
129 | 30.5M | c = charBuf; |
130 | 30.5M | charBuf = -1; |
131 | 30.5M | return c; |
132 | 30.5M | } |