Mercurial > cgi-bin > hgweb.cgi > curlyq
annotate curlers.py @ 11:1f5e471101b0
Deadwood; remove it.
| author | David Barts <n5jrn@me.com> |
|---|---|
| date | Fri, 27 Dec 2019 11:26:38 -0800 |
| parents | 397c178c5b98 |
| children | b2dab0667ec2 |
| rev | line source |
|---|---|
| 3 | 1 #!/usr/bin/env python3 |
| 2 # -*- coding: utf-8 -*- | |
| 3 | |
| 4 # Classes for curling both HTML and plain text. | |
| 5 | |
| 6 # I m p o r t s | |
| 7 | |
| 8 import os, sys | |
| 10 | 9 from runes import Workspace |
| 3 | 10 |
| 11 # V a r i a b l e s | |
| 12 | |
| 10 | 13 # Quote types, as rune values |
| 14 LSQUO = 0x2018 | |
| 15 APOS = RSQUO = 0x2019 | |
| 16 LDQUO = 0x201c | |
| 17 RDQUO = 0x201d | |
| 3 | 18 |
| 19 # Words that start with an apostrophe. Cribbed from Wordpress. | |
| 20 _ASTART = [ "'tain't", "'twere", "'twas", "'tis", "'twill", "'til", | |
| 21 "'bout", "'nuff", "'round", "'cause" , "'em" ] | |
| 22 | |
| 23 # HTML tags that enclose raw data | |
|
7
9df9ff8cecde
Undo that; ignoring <pre> is a sticky wicket.
David Barts <n5jrn@me.com>
parents:
6
diff
changeset
|
24 _RAW = set(["script", "style"]) |
| 3 | 25 |
| 26 # HTML block elements | |
| 27 _BLOCK = set([ | |
|
7
9df9ff8cecde
Undo that; ignoring <pre> is a sticky wicket.
David Barts <n5jrn@me.com>
parents:
6
diff
changeset
|
28 "address", "blockquote", "div", "dl", "fieldset", "form", "h1", |
|
9df9ff8cecde
Undo that; ignoring <pre> is a sticky wicket.
David Barts <n5jrn@me.com>
parents:
6
diff
changeset
|
29 "h2", "h3", "h4", "h5", "h6", "hr", "noscript", "ol", "p", "pre", |
|
9df9ff8cecde
Undo that; ignoring <pre> is a sticky wicket.
David Barts <n5jrn@me.com>
parents:
6
diff
changeset
|
30 "table", "ul" |
| 3 | 31 ]) |
| 32 | |
| 33 # F u n c t i o n s | |
| 34 | |
| 4 | 35 def uncurl(ws): |
| 36 """ | |
| 37 Makes all quotes in the workspace non-curly. | |
| 38 """ | |
| 39 for i in range(len(ws)): | |
| 40 ch = ws[i] | |
| 41 if ch in set([LDQUO, RDQUO]): | |
| 42 ws[i] = '"' | |
| 43 elif ch in set([LSQUO, RSQUO]): | |
| 44 ws[i] = "'" | |
| 45 | |
| 3 | 46 def _is_cockney(pos, ws): |
| 47 for i in _ASTART: | |
| 48 li = len(i) | |
| 49 if ws[pos:pos+li].lower() == i and not ws[pos+li].isalpha(): | |
| 50 return True | |
| 51 | |
| 52 # C l a s s e s | |
| 53 | |
| 54 class BaseCurler(): | |
| 55 def feed(self): | |
| 56 raise NotImplementedError() | |
| 57 | |
| 58 class TextCurler(BaseCurler): | |
| 59 """ | |
| 60 For processing plain text. Assumes the entire text is a block; it is | |
| 61 the responsibility of the caller to break the input into paragraphs. | |
| 62 """ | |
| 63 def __init__(self, workspace): | |
| 64 self.workspace = workspace | |
| 65 self._state = self._norm | |
| 66 self._pos = 0 | |
| 67 | |
| 68 def feed(self): | |
| 69 self._pos = 0 | |
| 70 self._state = self._norm | |
| 71 for self._pos in range(len(self.workspace)): | |
| 72 self._state() | |
| 73 | |
| 74 def _is_cockney(self): | |
| 75 pos = self._pos | |
| 76 ws = self.workspace | |
| 77 for i in _ASTART: | |
| 78 li = len(i) | |
| 79 print("comparing {0!r} and {1!r}\n".format(ws[pos:pos+li].lower(), i)) | |
| 80 if ws[pos:pos+li].lower() == i and not ws[pos+li].isalpha(): | |
| 81 return True | |
| 82 | |
| 83 def _norm(self): | |
| 84 pos = self._pos | |
| 85 ws = self.workspace | |
| 86 char = ws[pos] | |
| 87 if char == "\"": | |
| 88 # opening double quote | |
| 89 ws[pos] = LDQUO | |
| 90 self._state = self._seen_ld | |
| 91 elif char == "'": | |
| 92 # in this state, ' is always an apostrophe | |
| 93 ws[pos] = APOS | |
| 94 | |
| 95 def _seen_ld(self): | |
| 96 pos = self._pos | |
| 97 ws = self.workspace | |
| 98 char = ws[pos] | |
| 99 if char == "\"": | |
| 100 # closing double quote | |
| 101 ws[pos] = RDQUO | |
| 102 self._state = self._norm | |
| 103 elif char == "'": | |
| 104 if ws[pos-1].isalpha(): | |
| 105 # either an inter-word, or an end of word, apostrophe | |
| 106 ws[pos] = APOS | |
| 107 elif ws[pos+1].isdecimal() or _is_cockney(pos, ws): | |
| 108 # also an apostrophe | |
| 109 ws[pos] = APOS | |
| 110 else: | |
| 111 # opening single quote | |
| 112 ws[pos] = LSQUO | |
| 113 self._state = self._seen_ls | |
| 114 | |
| 115 def _seen_ls(self): | |
| 116 pos = self._pos | |
| 117 ws = self.workspace | |
| 118 if ws[pos] == "'": | |
| 119 if ws[pos-1].isalpha() and ws[pos+1].isalpha(): | |
| 120 # obvious apostrophe | |
| 121 ws[pos] = APOS | |
| 122 elif ws[pos+1].isdecimal() or _is_cockney(pos, ws): | |
| 123 # also an apostrophe | |
| 124 ws[pos] = APOS | |
| 125 elif ws[pos-1].isspace(): | |
| 126 # start of word apostrophe | |
| 127 ws[pos] = APOS | |
| 128 else: | |
| 129 # closing single quote | |
| 130 ws[pos] = RSQUO | |
| 131 self._state = self._seen_ld | |
| 132 | |
| 133 class HtmlCurler(BaseCurler): | |
| 134 """ | |
| 135 For processing HTML. Uses HTML block tags to delimit blocks. | |
| 136 """ | |
| 137 def __init__(self, workspace): | |
| 138 self.workspace = workspace | |
| 139 self._state = self._norm | |
| 140 self._pos = 0 | |
| 141 self._ltpos = 0 | |
| 142 self._endtag = None | |
| 143 self._ltstate = None | |
| 144 | |
| 145 def feed(self): | |
| 146 self._pos = 0 | |
| 147 self._state = self._norm | |
| 148 for self._pos in range(len(self.workspace)): | |
| 149 self._state() | |
| 150 | |
| 151 def _is_cockney(self): | |
| 152 pos = self._pos | |
| 153 ws = self.workspace | |
| 154 for i in _ASTART: | |
| 155 li = len(i) | |
| 156 print("comparing {0!r} and {1!r}\n".format(ws[pos:pos+li].lower(), i)) | |
| 157 if ws[pos:pos+li].lower() == i and not ws[pos+li].isalpha(): | |
| 158 return True | |
| 159 | |
| 160 def _goto_lt(self): | |
| 161 self._ltpos = self._pos | |
| 162 self._ltstate = self._state | |
| 163 self._state = self._seen_lt | |
| 164 | |
| 165 def _norm(self): | |
| 166 pos = self._pos | |
| 167 ws = self.workspace | |
| 168 char = ws[pos] | |
| 169 if char == "<": | |
| 170 self._goto_lt() | |
| 171 elif char == "\"": | |
| 172 # opening double quote | |
| 173 ws[pos] = LDQUO | |
| 174 self._state = self._seen_ld | |
| 175 elif char == "'": | |
| 176 # in this state, ' is always an apostrophe | |
| 177 ws[pos] = APOS | |
| 178 | |
| 179 def _gettag(self, start): | |
| 180 ws = self.workspace | |
| 181 end = start | |
| 182 while ws[end].isalnum(): | |
| 183 end += 1 | |
| 184 return ws[start:end].lower() | |
| 185 | |
| 186 def _seen_lt(self): | |
| 187 pos = self._pos | |
| 188 ws = self.workspace | |
| 189 if ws[pos] == ">": | |
| 190 start = self._ltpos + 1 | |
| 191 if ws[start] == '/': | |
| 192 if self._gettag(start + 1) in _BLOCK: | |
| 193 self._state = self._norm | |
| 194 else: | |
| 195 self._state = self._ltstate | |
| 196 else: | |
| 197 tag = self._gettag(start) | |
| 198 if tag in _BLOCK: | |
| 199 self._state = self._norm | |
| 200 elif tag in _RAW: | |
| 201 self._state = self._raw | |
| 202 self._endtag = "</" + tag | |
| 203 else: | |
| 204 self._state = self._ltstate | |
| 205 | |
| 206 def _raw(self): | |
| 207 pos = self._pos | |
| 208 ws = self.workspace | |
| 209 end = pos + len(self._endtag) | |
| 210 # only a matching end tag gets us out of the raw state | |
| 211 if ws[pos] == '<' and ws[pos:end].lower() == self._endtag and (not ws[end].isalnum()): | |
| 212 self._ltpos = pos | |
|
7
9df9ff8cecde
Undo that; ignoring <pre> is a sticky wicket.
David Barts <n5jrn@me.com>
parents:
6
diff
changeset
|
213 self._state = self._seen_lt |
| 3 | 214 |
| 215 def _seen_ld(self): | |
| 216 pos = self._pos | |
| 217 ws = self.workspace | |
| 218 char = ws[pos] | |
| 219 if char == "<": | |
| 220 self._goto_lt() | |
| 221 elif char == "\"": | |
| 222 # closing double quote | |
| 223 ws[pos] = RDQUO | |
| 224 self._state = self._norm | |
| 225 elif char == "'": | |
| 226 if ws[pos-1].isalpha(): | |
| 227 # either an inter-word, or an end of word, apostrophe | |
| 228 ws[pos] = APOS | |
| 229 elif ws[pos+1].isdecimal() or _is_cockney(pos, ws): | |
| 230 # also an apostrophe | |
| 231 ws[pos] = APOS | |
| 232 else: | |
| 233 # opening single quote | |
| 234 ws[pos] = LSQUO | |
| 235 self._state = self._seen_ls | |
| 236 | |
| 237 def _seen_ls(): | |
| 238 pos = self._pos | |
| 239 ws = self.workspace | |
| 240 char = ws[pos] | |
| 241 if char == "<": | |
| 242 self._goto_lt() | |
| 243 elif char == "'": | |
| 244 if ws[pos-1].isalpha() and ws[pos+1].isalpha(): | |
| 245 # obvious apostrophe | |
| 246 ws[pos] = APOS | |
| 247 elif ws[pos+1].isdecimal() or _is_cockney(pos, ws): | |
| 248 # also an apostrophe | |
| 249 ws[pos] = APOS | |
| 250 elif ws[pos-1].isspace(): | |
| 251 # start of word apostrophe | |
| 252 ws[pos] = APOS | |
| 253 else: | |
| 254 # closing single quote | |
| 255 ws[pos] = RSQUO | |
| 256 self._state = self._seen_ld |