too much, even for multilevel conversion.
(transcode_loop): use rb_econv_conv.
(econv_primitive_convert): ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@18610 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
transcode_output_followed_by_input.
* transcode.c (OUTPUT_FOLLOWED_BY_INPUT): new flag.
(transcode_restartable0): suspend when output followed by input if
OUTPUT_FOLLOWED_BY_INPUT is specified.
(trans_sweep): check OUTPUT_FOLLOWED_BY_INPUT.
(rb_trans_conv): support OUTPUT_FOLLOWED_BY_INPUT.
(econv_primitive_convert): return :output_followed_by_input for
transcode_output_followed_by_input.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@18608 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
(rb_cr_newline): new transcoder.
* transcode.c (trans_open_i): one more exra room for input newline
converter.
(rb_trans_open): crlf newline and cr newline implemented.
(Init_transcode): Encoding::Converter::CRLF_NEWLINE and
Encoding::Converter::LF_NEWLINE defined.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@18557 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
path.
(load_transcoder_entry): renamed from load_transcoder.
(load_transcoder): new function for loding transcoder by encoding
names.
(rb_transcoding_open_by_transcoder): extracted from
rb_transcoding_open.
(rb_transcoding_open): use load_transcoder and
rb_transcoding_open_by_transcoder.
(rb_trans_open_by_transcoder_entries): new function.
(trans_open_i): construct entries array.
(rb_trans_open): use rb_trans_open_by_transcoder_entries.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@18551 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* transcode.c (trans_open_i): just record from and to.
(rb_trans_open): load transcodings.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@18531 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* tool/transcode-tblgen.rb: 8bit byte of ASCII-8BIT is a valid
(but unique to ASCII-8BIT) character.
* transcode.c (rb_eConversionUndefined): new error.
(rb_eInvalidByteSequence): ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@18524 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
resetting a state of stateful encoding.
* enc/trans/iso2022.trans (rb_EUC_JP_to_ISO_2022_JP): specify
finish_eucjp_to_iso2022jp for resetstate_func.
* tool/transcode-tblgen.rb: specify NULL for resetstate_func.
* transcode.c (output_replacement_character): call resetstate_func
before appending the replacement character.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@18503 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
(rb_trans_elem_t): new type.
(rb_trans_t): new type.
* transcode.c (transcode_dispatch_cb): removed.
(transcode_dispatch): removed.
(rb_transcoding_result_t): moved to rb_trans_result_t in
transcode_data.h.
(transcode_restartable0): goto follow_info when FUNsi.
(rb_transcoding_open): use get_transcoder_entry.
(rb_trans_open): new function.
(rb_trans_conv): ditto.
(rb_trans_close): ditto.
(trans_open_i): ditto.
(trans_sweep): ditto.
(more_output_buffer): take rb_trans_t instead of rb_transcoding as
an argument.
(transcode_loop): take from_encoding and to_encoding instead of tr
as arguments. use rb_trans_open/rb_trans_conv/rb_trans_close.
(str_transcode): don't use transcode_dispatch.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@18498 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
(transcode_restartable): use PARTIAL_INPUT for converting buffered
input.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@18476 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* transcode.c (load_transcoder): extracted from transcode_dispatch_cb.
(rb_transcoding_result_t): renamed from transcode_result_t.
(rb_transcoding_open): new function.
(rb_transcoding_convert): ditto.
(rb_transcoding_close): ditto.
(transcode_loop): use rb_transcoding_open, rb_transcoding_convert
and rb_transcoding_close.
(str_transcode): don't need rb_transcoding.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@18474 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* transcode.c (transcode_restartable0): renamed from
transcode_restartable.
save input buffer into feed buffer if next character is started the
point before input buffer. for example, "\x00\xd8\x01" then "\x02"
in UTF-16LE. \x02 causes invalid and next character is started from
\x01.
(transcode_restartable): new function to call
transcode_restartable0. if feed buffer is not empty, convert it at
first.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@18467 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
as parameters.
(more_output_buffer): ditto.
(str_transcoding_resize): argument changed from rb_transcoding* to
VALUE.
(str_transcode): call transcode_loop with destination string and its
resize function.
* transcode_data.h (rb_transcoding): move ruby_string_dest and
flush_func to transcode_loop parameters.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@18458 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
(transcode_restartable): arguments changed to avoid *in_pos points
out of buffer by decreasing *in_pos.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@18455 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
(rb_transcoder): preprocessor and postprocessor field removed.
change arguments of func_ii, func_si, func_io and func_so.
new field "finish_func".
* tool/transcode-tblgen.rb: make FUNii, FUNsi and FUNio
generatable.
* transcode.c (transcoder_lib_table): removed.
(transcoder_table): change structure.
(transcoder_key): removed because the above structure change.
(make_transcoder_entry): new function.
(get_transcoder_entry): ditto.
(rb_register_transcoder): follow the structure change.
(declare_transcoder): ditto.
(transcode_search_path): new function for breadth first search to
find a list of converters.
(transcode_search_path_i): new function.
(transcode_dispatch_cb): ditto.
(transcode_dispatch): use transcode_search_path.
(transcode_loop): follow the argument change.
(str_transcode): preprocessor and postprocessor stuff removed.
* enc/trans/iso2022.erb.c: new file. ISO-2022-JP conversion
re-implemented.
* enc/trans/japanese.erb.c: ISO-2022-JP stuff removed.
nute(23:52:53)% head -40 ChangeLog
Thu Aug 7 23:43:11 2008 Tanaka Akira <akr@fsij.org>
* transcode_data.h (rb_transcoding): new field "stateful".
(rb_transcoder): preprocessor and postprocessor field removed.
change arguments of func_ii, func_si, func_io and func_so.
new field "finish_func".
* tool/transcode-tblgen.rb: make FUNii, FUNsi and FUNio
generatable.
* transcode.c (transcoder_lib_table): removed.
(transcoder_table): change structure.
(transcoder_key): removed because the above structure change.
(make_transcoder_entry): new function.
(get_transcoder_entry): ditto.
(rb_register_transcoder): follow the structure change.
(declare_transcoder): ditto.
(transcode_search_path): new function for breadth first search to
find a list of converters.
(transcode_search_path_i): new function.
(transcode_dispatch_cb): ditto.
(transcode_dispatch): use transcode_search_path.
(transcode_loop): follow the argument change.
(str_transcode): preprocessor and postprocessor stuff removed.
* enc/trans/iso2022.erb.c: new file. ISO-2022-JP conversion
re-implemented.
* enc/trans/japanese.erb.c: ISO-2022-JP stuff removed.
* enc/trans/utf_16_32.erb.c: follow argument change of FUNso.
[ruby-dev:35798]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@18419 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
string.c (rb_str_replace), transcode.c (transcode_dispatch): fixed
memory leaks. based on patches from shinichiro.h <shinichiro.hamaji
AT gmail.com> at [ruby-dev:35751].
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@18341 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
character when convert to Unicode.
* test/ruby/test_transcode.rb (test_unicode_public_review_issue_121):
rename from test_public_review_issue_121.
* test/ruby/test_transcode.rb (test_unicode_public_review_issue_121):
enable option2.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@18294 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
