Class: Docurium::CParser
- Inherits:
-
Object
- Object
- Docurium::CParser
- Defined in:
- lib/docurium/cparser.rb
Constant Summary collapse
- DETECTORS =
Array of detectors to execute in order
%w(define enum struct typedef function catchall)
Instance Method Summary collapse
-
#cleanup_comment(comment) ⇒ Object
Remove common prefix of all lines in comment.
-
#detect_catchall(d) ⇒ Object
Match otherwise unrecognized commented blocks.
-
#detect_define(d) ⇒ Object
Match #define A(B) or #define A and convert a series of simple #defines into an enum.
-
#detect_enum(d) ⇒ Object
Match enum {} (and possibly a typedef thereof).
-
#detect_function(d) ⇒ Object
Match function prototypes or inline function declarations.
-
#detect_struct(d) ⇒ Object
Match struct {} (and typedef thereof) or opaque struct typedef.
-
#detect_typedef(d) ⇒ Object
Match other typedefs, checking explicitly for function pointers but otherwise just trying to extract a name as simply as possible.
-
#join_define(text) ⇒ Object
Take a multi-line #define and join into a simple definition.
-
#parse_declaration_block(d) ⇒ Object
Given a commented chunk of file, try to parse it.
-
#parse_define(d) ⇒ Object
Process #define A …
-
#parse_enum(d) ⇒ Object
Process enum definitions.
-
#parse_file(d) ⇒ Object
Process comment blocks that are only associated with the whole file.
-
#parse_fnptr(d) ⇒ Object
Process function pointer typedef definition.
-
#parse_function(d) ⇒ Object
Process function prototype and comments.
-
#parse_macro(d) ⇒ Object
Process #define A(B) macros.
-
#parse_struct(d) ⇒ Object
Process struct definition.
-
#parse_text(filename, content) ⇒ Object
Parse a chunk of text as a header file.
-
#parse_typedef(d) ⇒ Object
Process typedef definition.
-
#shortest_common_prefix(arr) ⇒ Object
Find the shortest common prefix of an array of strings.
Instance Method Details
#cleanup_comment(comment) ⇒ Object
Remove common prefix of all lines in comment. Otherwise tries to preserve formatting in case it is relevant.
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 |
# File 'lib/docurium/cparser.rb', line 6 def cleanup_comment(comment) return "" unless comment lines = 0 prefixed = 0 shortest = nil compacted = comment.sub(/^\n+/,"").sub(/\n\s*$/, "\n") compacted.split(/\n/).each do |line| lines += 1 if line =~ /^\s*\*\s*$/ || line =~ /^\s*$/ # don't count length of blank lines or lines with just a " * " on # them towards shortest common prefix prefixed += 1 shortest = line if shortest.nil? elsif line =~ /^(\s*\*\s*)/ prefixed += 1 shortest = $1 if shortest.nil? || shortest.length > $1.length end end if shortest =~ /\s$/ shortest = Regexp.quote(shortest.chop) + "[ \t]" elsif shortest shortest = Regexp.quote(shortest) end if lines == prefixed && !shortest.nil? && shortest.length > 0 if shortest =~ /\*/ return comment.gsub(/^#{shortest}/, "").gsub(/^\s*\*\s*\n/, "\n") else return comment.gsub(/^#{shortest}/, "") end else return comment end end |
#detect_catchall(d) ⇒ Object
Match otherwise unrecognized commented blocks
311 312 313 314 |
# File 'lib/docurium/cparser.rb', line 311 def detect_catchall(d) d[:type] = :file d[:decl] = "" end |
#detect_define(d) ⇒ Object
Match #define A(B) or #define A and convert a series of simple #defines into an enum
54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 |
# File 'lib/docurium/cparser.rb', line 54 def detect_define(d) if d[:body] =~ /\A\#\s*define\s+((\w+)\([^\)]+\))/ d[:type] = :macro d[:decl] = $1.strip d[:name] = $2 d[:tdef] = nil elsif d[:body] =~ /\A\#\s*define\s+(\w+)/ names = [] d[:body].scan(/\#\s*define\s+(\w+)/) { |m| names << m[0].to_s } d[:tdef] = nil names.uniq! if names.length == 1 d[:type] = :define d[:decl] = names[0] d[:name] = names[0] elsif names.length > 1 d[:type] = :enum d[:decl] = names d[:name] = shortest_common_prefix(names) d[:name].sub!(/_*$/, '') end end end |
#detect_enum(d) ⇒ Object
Match enum {} (and possibly a typedef thereof)
106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 |
# File 'lib/docurium/cparser.rb', line 106 def detect_enum(d) if d[:body] =~ /\A(typedef)?\s*enum\s*\{([^\}]+)\}\s*([^;]+)?;/i typedef, values, name = $1, $2, $3 d[:type] = :enum d[:decl] = values.strip.split(/\s*,\s*/).map do |v| v.split(/\s*=\s*/)[0].strip end if typedef.nil? d[:name] = shortest_common_prefix(d[:decl]) d[:name].sub!(/_*$/, '') # Using the common prefix for grouping enum values is a little # overly aggressive in some cases. If we ended up with too short # a prefix or a prefix which is too generic, then skip it. d[:name] = nil unless d[:name].scan('_').length > 1 else d[:name] = name end d[:tdef] = typedef end end |
#detect_function(d) ⇒ Object
Match function prototypes or inline function declarations
191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 |
# File 'lib/docurium/cparser.rb', line 191 def detect_function(d) if d[:body] =~ /[;\{]/ d[:type] = :file d[:decl] = "" proto = d[:body].split(/[;\{]/, 2).first.strip if proto[-1] == ?) (proto.length - 1).downto(0) do |p| tail = proto[p .. -1] if tail.count(")") == tail.count("(") if proto[0..p] =~ /(\w+)\(\z/ d[:name] = $1 d[:type] = :function d[:decl] = proto end break end end end end end |
#detect_struct(d) ⇒ Object
Match struct {} (and typedef thereof) or opaque struct typedef
138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 |
# File 'lib/docurium/cparser.rb', line 138 def detect_struct(d) if d[:body] =~ /\A(typedef)?\s*struct\s*(\w+)?\s*\{([^\}]+)\}\s*([^;]+)?/i typedef, name1, fields, name2 = $1, $2, $3, $4 d[:type] = :struct d[:name] = typedef.nil? ? name1 : name2; d[:tdef] = typedef d[:decl] = fields.strip.split(/\s*\;\s*/).map do |x| x.strip.gsub(/\s+/, " ").gsub(/\(\s+/,"(") end elsif d[:body] =~ /\A(typedef)\s+struct\s+\w+\s+(\w+)/ d[:type] = :struct d[:decl] = "" d[:name] = $2 d[:tdef] = $1 end end |
#detect_typedef(d) ⇒ Object
Match other typedefs, checking explicitly for function pointers but otherwise just trying to extract a name as simply as possible.
167 168 169 170 171 172 173 174 175 176 177 178 |
# File 'lib/docurium/cparser.rb', line 167 def detect_typedef(d) if d[:body] =~ /\Atypedef\s+([^;]+);/ d[:decl] = $1.strip if d[:decl] =~ /\S+\s+\(\*([^\)]+)\)\(/ d[:type] = :fnptr d[:name] = $1 else d[:type] = :typedef d[:name] = d[:decl].split(/\s+/).last end end end |
#join_define(text) ⇒ Object
Take a multi-line #define and join into a simple definition
79 80 81 82 83 84 85 86 87 |
# File 'lib/docurium/cparser.rb', line 79 def join_define(text) text = text.split("\n\n", 2).first || "" # Ruby 1.8 does not support negative lookbehind regex so let's # get the joined macro definition a slightly more awkward way text.split(/\s*\n\s*/).inject("\\") do |val, line| (val[-1] == ?\\) ? val = val.chop.strip + " " + line : val end.strip.gsub(/^\s*\\*\s*/, '') end |
#parse_declaration_block(d) ⇒ Object
Given a commented chunk of file, try to parse it.
334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 |
# File 'lib/docurium/cparser.rb', line 334 def parse_declaration_block(d) # skip uncommented declarations return unless d[:rawComments].length > 0 # remove inline comments in declaration while comment = d[:body].index("/*") do end_comment = d[:body].index("*/", comment) d[:body].slice!(comment, end_comment - comment + 2) end # if there are multiple #ifdef'ed declarations, we'll just # strip out the #if/#ifdef and use the first one d[:body].sub!(/[^\n]+\n/, '') if d[:body] =~ /\A\#\s*if/ # try detectors until one assigns a :type to the declaration # it's going to be one of: # - :define -> #defines + convert a series of simple ones to :enum # - :enum -> (typedef)? enum { ... }; # - :struct -> (typedef)? struct { ... }; # - :fnptr -> typedef x (*fn)(...); # - :typedef -> typedef x y; (not enum, struct, fnptr) # - :function -> rval something(like this); # - :file -> everything else goes to "file" scope DETECTORS.find { |p| method("detect_#{p}").call(d); d.has_key?(:type) } # if we detected something, call a parser for that type of thing method("parse_#{d[:type]}").call(d) if d[:type] end |
#parse_define(d) ⇒ Object
Process #define A … macros
98 99 100 101 102 103 |
# File 'lib/docurium/cparser.rb', line 98 def parse_define(d) if d[:body] =~ /define\s+#{Regexp.quote(d[:name])}[ \t]*(.*)/ d[:value] = join_define($1) end d[:comments] = d[:rawComments].strip end |
#parse_enum(d) ⇒ Object
Process enum definitions
128 129 130 131 132 133 134 135 |
# File 'lib/docurium/cparser.rb', line 128 def parse_enum(d) if d[:decl].respond_to? :map d[:block] = d[:decl].map { |v| v.strip }.join("\n") else d[:block] = d[:decl] end d[:comments] = d[:rawComments].strip end |
#parse_file(d) ⇒ Object
Process comment blocks that are only associated with the whole file.
317 318 319 320 321 322 323 324 325 326 327 328 |
# File 'lib/docurium/cparser.rb', line 317 def parse_file(d) m = [] d[:brief] = m[1] if m = /@brief (.*?)$/.match(d[:rawComments]) d[:defgroup] = m[1] if m = /@defgroup (.*?)$/.match(d[:rawComments]) d[:ingroup] = m[1] if m = /@ingroup (.*?)$/.match(d[:rawComments]) comments = d[:rawComments].gsub(/^@.*$/, '').strip + "\n" if d[:comments] d[:comments] = d[:comments].strip + "\n\n" + comments else d[:comments] = comments end end |
#parse_fnptr(d) ⇒ Object
Process function pointer typedef definition
186 187 188 |
# File 'lib/docurium/cparser.rb', line 186 def parse_fnptr(d) d[:comments] = d[:rawComments].strip end |
#parse_function(d) ⇒ Object
Process function prototype and comments
214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 |
# File 'lib/docurium/cparser.rb', line 214 def parse_function(d) d[:args] = [] rval, argline = d[:decl].split(/\s*#{Regexp.quote(d[:name])}\(\s*/, 2) # clean up rval if it is like "extern static int" or "GIT_EXTERN(int)" while rval =~ /[A-Za-z0-9_]+\(([^\)]+)\)$/i rval = $1 end rval.gsub!(/extern|static/, '') rval.strip! d[:return] = { :type => rval } # we removed the opening parenthesis, which this is expecting argline = '(' + argline # clean up argline argline = argline.slice(1..-2) while argline[0] == ?( && argline[-1] == ?) d[:argline] = argline.strip d[:args] = [] left = 0 # parse argline (0 .. argline.length).each do |i| next unless argline[i] == ?, || argline.length == i s = argline.slice(left .. i) next unless s.count("(") == s.count(")") s.chop! if argline[i] == ?, s.strip! if s =~ /\(\s*\*\s*(\w+)\s*\)\s*\(/ argname = $1 d[:args] << { :name => argname, :type => s.sub(/\s*#{Regexp.quote(argname)}\s*/, '').strip } elsif s =~ /\W(\w+)$/ argname = $1 d[:args] << { :name => argname, :type => s[0 ... - argname.length].strip, } else # argline is probably something like "(void)" end left = i + 1 end # parse comments if d[:rawComments] =~ /\@(param|return)/i d[:args].each do |arg| param_comment = /\@param\s+#{Regexp.quote(arg[:name])}/.match(d[:rawComments]) if param_comment after = param_comment.post_match end_comment = after.index(/(?:@param|@return|\Z)/) arg[:comment] = after[0 ... end_comment].strip.gsub(/\s+/, ' ') end end return_comment = /\@return\s+/.match(d[:rawComments]) if return_comment after = return_comment.post_match d[:return][:comment] = after[0 ... after.index(/(?:@param|\Z)/)].strip.gsub(/\s+/, ' ') end else # support for TomDoc params end # add in inline parameter comments if d[:inlines] # array of [param line]/[comment] pairs d[:inlines].each do |inl| d[:args].find do |arg| if inl[0] =~ /\b#{Regexp.quote(arg[:name])}$/ arg[:comment] += "\n#{inl[1]}" end end end end # generate function signature d[:sig] = d[:args].map { |a| a[:type].to_s }.join('::') # pull off function description if d[:rawComments] =~ /^\s*(public|internal|deprecated):/i # support for TomDoc description else desc, comments = d[:rawComments].split("\n\n", 2) d[:description] = desc.strip d[:comments] = comments || "" params_start = d[:comments].index(/\s?\@(?:param|return)/) d[:comments] = d[:comments].slice(0, params_start) if params_start end end |
#parse_macro(d) ⇒ Object
Process #define A(B) macros
90 91 92 93 94 95 |
# File 'lib/docurium/cparser.rb', line 90 def parse_macro(d) if d[:body] =~ /define\s+#{Regexp.quote(d[:name])}\([^\)]*\)[ \t]*(.*)/m d[:value] = join_define($1) end d[:comments] = d[:rawComments].strip end |
#parse_struct(d) ⇒ Object
Process struct definition
156 157 158 159 160 161 162 163 |
# File 'lib/docurium/cparser.rb', line 156 def parse_struct(d) if d[:decl].respond_to? :map d[:block] = d[:decl].map { |v| v.strip }.join("\n") else d[:block] = d[:decl] end d[:comments] = d[:rawComments].strip end |
#parse_text(filename, content) ⇒ Object
Parse a chunk of text as a header file
364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 |
# File 'lib/docurium/cparser.rb', line 364 def parse_text(filename, content) # break into comments and non-comments with line numbers content = "/** */" + content if content[0..2] != "/**" recs = [] lineno = 1 openblock = false content.split(/\/\*\*/).each do |chunk| c, b = chunk.split(/[ \t]*\*\//, 2) next unless c || b lineno += c.scan("\n").length if c # special handling for /**< ... */ inline comments or # for /** ... */ inside an open block if openblock || c[0] == ?< c = c.sub(/^</, '').strip so_far = recs[-1][:body] last_line = so_far[ so_far.rindex("\n")+1 .. -1 ].strip.chomp(",").chomp(";") if last_line.empty? && b =~ /^([^;]+)\;/ # apply to this line instead last_line = $1.strip.chomp(",").chomp(";") end if !last_line.empty? recs[-1][:inlines] ||= [] recs[-1][:inlines] << [ last_line, c ] if b recs[-1][:body] += b lineno += b.scan("\n").length openblock = false if b =~ /\}/ end next end end # make comment have a uniform " *" prefix if needed if c !~ /\A[ \t]*\n/ && c =~ /^(\s*\*)/ c = $1 + c end # check for unterminated { brace (to handle inline comments later) openblock = true if b =~ /\{[^\}]+\Z/ recs << { :file => filename, :line => lineno + (b.start_with?("\n") ? 1 : 0), :body => b, :rawComments => cleanup_comment(c), } lineno += b.scan("\n").length if b end # try parsers on each chunk of commented header recs.each do |r| r[:body].strip! r[:rawComments].strip! r[:lineto] = r[:line] + r[:body].scan("\n").length parse_declaration_block(r) end recs end |
#parse_typedef(d) ⇒ Object
Process typedef definition
181 182 183 |
# File 'lib/docurium/cparser.rb', line 181 def parse_typedef(d) d[:comments] = d[:rawComments].strip end |
#shortest_common_prefix(arr) ⇒ Object
Find the shortest common prefix of an array of strings
46 47 48 49 50 |
# File 'lib/docurium/cparser.rb', line 46 def shortest_common_prefix(arr) arr.inject do |pfx,str| pfx = pfx.chop while pfx != str[0...pfx.length]; pfx end end |