Guido van Rossum | c636014 | 1990-10-13 19:23:40 +0000 | [diff] [blame] | 1 | # Module 'parser' |
| 2 | # |
| 3 | # Parse S-expressions output by the Panel Editor |
| 4 | # (which is written in Scheme so it can't help writing S-expressions). |
| 5 | # |
| 6 | # See notes at end of file. |
| 7 | |
| 8 | |
| 9 | whitespace = ' \t\n' |
| 10 | operators = '()\'' |
| 11 | separators = operators + whitespace + ';' + '"' |
| 12 | |
| 13 | |
| 14 | # Tokenize a string. |
| 15 | # Return a list of tokens (strings). |
| 16 | # |
| 17 | def tokenize_string(s): |
| 18 | tokens = [] |
| 19 | while s: |
| 20 | c = s[:1] |
| 21 | if c in whitespace: |
| 22 | s = s[1:] |
Guido van Rossum | bdfcfcc | 1992-01-01 19:35:13 +0000 | [diff] [blame] | 23 | elif c == ';': |
Guido van Rossum | c636014 | 1990-10-13 19:23:40 +0000 | [diff] [blame] | 24 | s = '' |
Guido van Rossum | bdfcfcc | 1992-01-01 19:35:13 +0000 | [diff] [blame] | 25 | elif c == '"': |
Guido van Rossum | c636014 | 1990-10-13 19:23:40 +0000 | [diff] [blame] | 26 | n = len(s) |
| 27 | i = 1 |
| 28 | while i < n: |
| 29 | c = s[i] |
| 30 | i = i+1 |
Guido van Rossum | bdfcfcc | 1992-01-01 19:35:13 +0000 | [diff] [blame] | 31 | if c == '"': break |
| 32 | if c == '\\': i = i+1 |
Guido van Rossum | c636014 | 1990-10-13 19:23:40 +0000 | [diff] [blame] | 33 | tokens.append(s[:i]) |
| 34 | s = s[i:] |
| 35 | elif c in operators: |
| 36 | tokens.append(c) |
| 37 | s = s[1:] |
| 38 | else: |
| 39 | n = len(s) |
| 40 | i = 1 |
| 41 | while i < n: |
| 42 | if s[i] in separators: break |
| 43 | i = i+1 |
| 44 | tokens.append(s[:i]) |
| 45 | s = s[i:] |
| 46 | return tokens |
| 47 | |
| 48 | |
| 49 | # Tokenize a whole file (given as file object, not as file name). |
| 50 | # Return a list of tokens (strings). |
| 51 | # |
| 52 | def tokenize_file(fp): |
| 53 | tokens = [] |
| 54 | while 1: |
| 55 | line = fp.readline() |
| 56 | if not line: break |
| 57 | tokens = tokens + tokenize_string(line) |
| 58 | return tokens |
| 59 | |
| 60 | |
| 61 | # Exception raised by parse_exr. |
| 62 | # |
| 63 | syntax_error = 'syntax error' |
| 64 | |
| 65 | |
| 66 | # Parse an S-expression. |
| 67 | # Input is a list of tokens as returned by tokenize_*(). |
| 68 | # Return a pair (expr, tokens) |
| 69 | # where expr is a list representing the s-expression, |
| 70 | # and tokens contains the remaining tokens. |
| 71 | # May raise syntax_error. |
| 72 | # |
| 73 | def parse_expr(tokens): |
| 74 | if (not tokens) or tokens[0] <> '(': |
| 75 | raise syntax_error, 'expected "("' |
| 76 | tokens = tokens[1:] |
| 77 | expr = [] |
| 78 | while 1: |
| 79 | if not tokens: |
| 80 | raise syntax_error, 'missing ")"' |
Guido van Rossum | bdfcfcc | 1992-01-01 19:35:13 +0000 | [diff] [blame] | 81 | if tokens[0] == ')': |
Guido van Rossum | c636014 | 1990-10-13 19:23:40 +0000 | [diff] [blame] | 82 | return expr, tokens[1:] |
Guido van Rossum | bdfcfcc | 1992-01-01 19:35:13 +0000 | [diff] [blame] | 83 | elif tokens[0] == '(': |
Guido van Rossum | c636014 | 1990-10-13 19:23:40 +0000 | [diff] [blame] | 84 | subexpr, tokens = parse_expr(tokens) |
| 85 | expr.append(subexpr) |
| 86 | else: |
| 87 | expr.append(tokens[0]) |
| 88 | tokens = tokens[1:] |
| 89 | |
| 90 | |
| 91 | # Parse a file (given as file object, not as file name). |
| 92 | # Return a list of parsed S-expressions found at the top level. |
| 93 | # |
| 94 | def parse_file(fp): |
| 95 | tokens = tokenize_file(fp) |
| 96 | exprlist = [] |
| 97 | while tokens: |
| 98 | expr, tokens = parse_expr(tokens) |
| 99 | exprlist.append(expr) |
| 100 | return exprlist |
| 101 | |
| 102 | |
| 103 | # EXAMPLE: |
| 104 | # |
| 105 | # The input |
| 106 | # '(hip (hop hur-ray))' |
| 107 | # |
| 108 | # passed to tokenize_string() returns the token list |
| 109 | # ['(', 'hip', '(', 'hop', 'hur-ray', ')', ')'] |
| 110 | # |
| 111 | # When this is passed to parse_expr() it returns the expression |
| 112 | # ['hip', ['hop', 'hur-ray']] |
| 113 | # plus an empty token list (because there are no tokens left. |
| 114 | # |
| 115 | # When a file containing the example is passed to parse_file() it returns |
| 116 | # a list whose only element is the output of parse_expr() above: |
| 117 | # [['hip', ['hop', 'hur-ray']]] |
| 118 | |
| 119 | |
| 120 | # TOKENIZING: |
| 121 | # |
| 122 | # Comments start with semicolon (;) and continue till the end of the line. |
| 123 | # |
| 124 | # Tokens are separated by whitespace, except the following characters |
| 125 | # always form a separate token (outside strings): |
| 126 | # ( ) ' |
| 127 | # Strings are enclosed in double quotes (") and backslash (\) is used |
| 128 | # as escape character in strings. |