1 # -*- coding: iso-8859-1 -*- 2 """ 3 MoinMoin - ImprovedTableParser library 4 5 @copyright: 2012 by Paul Boddie <paul@boddie.org.uk> 6 @license: GNU GPL (v2 or later), see COPYING.txt for details. 7 """ 8 9 from MoinMoin import wikiutil 10 from shlex import shlex 11 from StringIO import StringIO 12 from MoinSupport import * 13 import re 14 15 # Regular expressions. 16 17 syntax = { 18 # For section markers. 19 "markers" : (r"^\s*(?P<n>\\+)(?P<b>{|})(?P=n)(?P=b)(?P=n)(?P=b)", re.MULTILINE), 20 "marker" : (r"(\\+)", 0), 21 22 # At start of line: 23 "sections" : (r"(^\s*{{{.*?^\s*}}})", re.MULTILINE | re.DOTALL), # {{{ ... }}} 24 "rows" : (r"^==", re.MULTILINE), # == 25 26 # Within text: 27 "columns" : (r"\|\|[ \t]*", 0), # || ws-excl-nl 28 29 # At start of column text: 30 "column" : (r"^\s*<(.*?)>\s*(.*)", re.DOTALL), # ws < attributes > ws 31 } 32 33 patterns = {} 34 for name, (value, flags) in syntax.items(): 35 patterns[name] = re.compile(value, re.UNICODE | flags) 36 37 # Other regular expressions. 38 39 leading_number_regexp = re.compile(r"\d*") 40 41 # Functions. 42 43 def parse(s): 44 45 "Parse 's', returning a table definition." 46 47 s = replaceMarkers(s) 48 49 table_attrs = {} 50 rows = [] 51 52 # The following will be redefined upon the construction of the first row. 53 54 row_attrs = {} 55 columns = [] 56 57 # Process exposed text and sections. 58 59 exposed = True 60 61 # Initially, start a new row. 62 63 row_continued = False 64 65 for region in patterns["sections"].split(s): 66 67 # Only look for table features in exposed text. 68 69 if exposed: 70 71 # Extract each row from the definition. 72 73 for row_text in patterns["rows"].split(region): 74 75 # Only create a new row when a boundary has been found. 76 77 if not row_continued: 78 if columns: 79 extractAttributes(columns[0][0], row_attrs, table_attrs) 80 81 row_attrs = {} 82 columns = [] 83 rows.append((row_attrs, columns)) 84 column_continued = False 85 86 # Extract each column from the row. 87 88 for text in patterns["columns"].split(row_text): 89 90 # Only create a new column when a boundary has been found. 91 92 if not column_continued: 93 94 # Extract the attribute and text sections. 95 96 match = patterns["column"].search(text) 97 if match: 98 attribute_text, text = match.groups() 99 columns.append([parseAttributes(attribute_text, True), text]) 100 else: 101 columns.append([{}, text]) 102 103 else: 104 columns[-1][1] += text 105 106 # Permit columns immediately following this one. 107 108 column_continued = False 109 110 # Permit a continuation of the current column. 111 112 column_continued = True 113 114 # Permit rows immediately following this one. 115 116 row_continued = False 117 118 # Permit a continuation if the current row. 119 120 row_continued = True 121 122 # Write any section into the current column. 123 124 else: 125 columns[-1][1] += region 126 127 exposed = not exposed 128 129 if columns: 130 extractAttributes(columns[0][0], row_attrs, table_attrs) 131 132 return table_attrs, rows 133 134 def extractAttributes(attrs, row_attrs, table_attrs): 135 136 """ 137 Extract row- and table-level attributes from 'attrs', storing them in 138 'row_attrs' and 'table_attrs' respectively. 139 """ 140 141 for name, value in attrs.items(): 142 if name.startswith("row") and name != "rowspan": 143 row_attrs[name] = value 144 del attrs[name] 145 elif name.startswith("table"): 146 table_attrs[name] = value 147 del attrs[name] 148 149 def replaceMarkers(s): 150 151 "Convert the section notation in 's'." 152 153 l = [] 154 last = 0 155 156 # Get each marker and convert it. 157 158 for match in patterns["markers"].finditer(s): 159 start, stop = match.span() 160 l.append(s[last:start]) 161 162 # Convert the marker. 163 164 marker = [] 165 brace = True 166 for text in patterns["marker"].split(match.group()): 167 if brace: 168 marker.append(text) 169 else: 170 marker.append(text[:-1]) 171 brace = not brace 172 173 l.append("".join(marker)) 174 last = stop 175 else: 176 l.append(s[last:]) 177 178 return "".join(l) 179 180 def parseAttributes(s, escape=True): 181 182 """ 183 Parse the table attributes string 's', returning a mapping of names to 184 values. If 'escape' is set to a true value, the attributes will be suitable 185 for use with the formatter API. If 'escape' is set to a false value, the 186 attributes will have any quoting removed. 187 """ 188 189 attrs = {} 190 f = StringIO(s) 191 name = None 192 need_value = False 193 194 for token in shlex(f): 195 196 # Capture the name if needed. 197 198 if name is None: 199 name = escape and wikiutil.escape(token) or strip_token(token) 200 201 # Detect either an equals sign or another name. 202 203 elif not need_value: 204 if token == "=": 205 need_value = True 206 else: 207 attrs[name.lower()] = escape and "true" or True 208 name = wikiutil.escape(token) 209 210 # Otherwise, capture a value. 211 212 else: 213 # Quoting of attributes done similarly to wikiutil.parseAttributes. 214 215 if token: 216 if escape: 217 if token[0] in ("'", '"'): 218 token = wikiutil.escape(token) 219 else: 220 token = '"%s"' % wikiutil.escape(token, 1) 221 else: 222 token = strip_token(token) 223 224 attrs[name.lower()] = token 225 name = None 226 need_value = False 227 228 return attrs 229 230 def strip_token(token): 231 232 "Return the given 'token' stripped of quoting." 233 234 if token[0] in ("'", '"') and token[-1] == token[0]: 235 return token[1:-1] 236 else: 237 return token 238 239 # Formatting of embedded content. 240 # NOTE: Borrowed from EventAggregator. 241 242 def getParserClass(request, format): 243 244 """ 245 Return a parser class using the 'request' for the given 'format', returning 246 a plain text parser if no parser can be found for the specified 'format'. 247 """ 248 249 try: 250 return wikiutil.searchAndImportPlugin(request.cfg, "parser", format or "plain") 251 except wikiutil.PluginMissingError: 252 return wikiutil.searchAndImportPlugin(request.cfg, "parser", "plain") 253 254 def formatText(text, request, fmt): 255 256 "Format the given 'text' using the specified 'request' and formatter 'fmt'." 257 258 parser_cls = getParserClass(request, request.page.pi["format"]) 259 parser = parser_cls(text, request, line_anchors=False) 260 return request.redirectedOutput(parser.format, fmt, inhibit_p=True) 261 262 # Sorting utilities. 263 264 def get_sort_columns(s, start=0): 265 266 """ 267 Split the comma-separated string 's', extracting the column specifications 268 of the form <column>["n"] where the prefix "n" indicates an optional 269 numeric conversion for that column. Column indexes start from the specified 270 'start' value (defaulting to 0). 271 """ 272 273 sort_columns = [] 274 for column_spec in s.split(","): 275 column_spec = column_spec.strip() 276 277 ascending = True 278 if column_spec.endswith("d"): 279 column_spec = column_spec[:-1] 280 ascending = False 281 282 # Extract the conversion indicator and column index. 283 284 if column_spec.endswith("n"): 285 column = column_spec[:-1] 286 fn = to_number 287 else: 288 column = column_spec 289 fn = str 290 291 # Ignore badly-specified columns. 292 293 try: 294 sort_columns.append((max(0, int(column) - start), fn, ascending)) 295 except ValueError: 296 pass 297 298 return sort_columns 299 300 def to_number(s): 301 302 "Convert 's' to a number, discarding any non-numeric trailing data." 303 304 match = leading_number_regexp.match(s) 305 if match: 306 return int(match.group()) 307 else: 308 raise ValueError, s 309 310 class Sorter: 311 312 "A sorting helper class." 313 314 def __init__(self, sort_columns): 315 self.sort_columns = sort_columns 316 317 def __call__(self, row1, row2): 318 row_attrs1, columns1 = row1 319 row_attrs2, columns2 = row2 320 321 # Apply the conversions to each column, comparing the results. 322 323 for column, fn, ascending in self.sort_columns: 324 column_attrs1, text1 = columns1[column] 325 column_attrs2, text2 = columns2[column] 326 327 # Ignore a column when a conversion is not possible. 328 329 try: 330 text1 = fn(text1) 331 text2 = fn(text2) 332 result = cmp(text1, text2) 333 334 # Where the columns differ, return a result observing the sense 335 # (ascending or descending) of the comparison for the column. 336 337 if result != 0: 338 return ascending and result or -result 339 340 except ValueError: 341 pass 342 343 return 0 344 345 # Common formatting functions. 346 347 def formatTable(text, request, fmt, attrs=None): 348 349 """ 350 Format the given 'text' using the specified 'request' and formatter 'fmt'. 351 The optional 'attrs' can be used to control the presentation of the table. 352 """ 353 354 # Parse the table region. 355 356 table_attrs, table = parse(text) 357 358 # Override any region arguments with request parameters. 359 360 table_name = attrs.get("name") 361 sortcolumns = table_name and getQualifiedParameter(request, table_name, "sortcolumns") or attrs.get("sortcolumns") 362 363 # Sort the rows according to the values in each of the specified columns. 364 365 if sortcolumns: 366 data_start = int(attrs.get("headers", "1")) 367 headers = table[:data_start] 368 data = table[data_start:] 369 370 # Get the sort columns using Unix sort-like notation. 371 372 sorter = Sorter(get_sort_columns(sortcolumns)) 373 data.sort(cmp=sorter) 374 375 table = headers + data 376 377 # Write the table. 378 379 request.write(fmt.table(1, table_attrs)) 380 381 for row_attrs, columns in table: 382 request.write(fmt.table_row(1, row_attrs)) 383 384 for column_attrs, column_text in columns: 385 request.write(fmt.table_cell(1, column_attrs)) 386 request.write(formatText(column_text, request, fmt)) 387 request.write(fmt.table_cell(0)) 388 389 request.write(fmt.table_row(0)) 390 391 request.write(fmt.table(0)) 392 393 # vim: tabstop=4 expandtab shiftwidth=4