diff options
author | Sutou Kouhei <[email protected]> | 2021-12-24 10:44:11 +0900 |
---|---|---|
committer | Sutou Kouhei <[email protected]> | 2021-12-24 14:35:33 +0900 |
commit | 22ef4f6445376b992b2725124594dad1c77a185e (patch) | |
tree | 31c624339ba0f33258c6843e7b6cbc7d5bb00c76 /lib/csv/parser.rb | |
parent | 002ce9f515d984aa0fe43c51448b36ea3651a395 (diff) |
[ruby/csv] Revert "parser: fix a keep bug that some texts may be dropped unexpectedly"
This reverts commit https://github.com/ruby/csv/commit/5c6523da0a61.
This introduces another pbolem. We should try again later.
https://github.com/ruby/csv/commit/43a1d6fff1
Notes
Notes:
Merged: https://github.com/ruby/ruby/pull/5336
Diffstat (limited to 'lib/csv/parser.rb')
-rw-r--r-- | lib/csv/parser.rb | 85 |
1 files changed, 34 insertions, 51 deletions
diff --git a/lib/csv/parser.rb b/lib/csv/parser.rb index 78ff363d69..7e943acf21 100644 --- a/lib/csv/parser.rb +++ b/lib/csv/parser.rb @@ -167,7 +167,6 @@ class CSV end def keep_start - adjust_last_keep @keeps.push([@scanner.pos, nil]) end @@ -198,17 +197,7 @@ class CSV end def keep_drop - _, buffer = @keeps.pop - return unless buffer - - last_keep = @keeps.last - return unless last_keep - - if last_keep[1] - last_keep[1] << buffer - else - last_keep[1] = buffer - end + @keeps.pop end def rest @@ -216,30 +205,24 @@ class CSV end private - def adjust_last_keep - keep = @keeps.last - return if keep.nil? - - keep_start = keep[0] - return if @scanner.pos == keep_start - - string = @scanner.string - keep_data = string.byteslice(keep_start, @scanner.pos - keep_start) - if keep_data - keep_buffer = keep[1] - if keep_buffer - keep_buffer << keep_data - else - keep[1] = keep_data.dup - end - end - keep[0] = 0 - end - def read_chunk return false if @last_scanner - adjust_last_keep + unless @keeps.empty? + keep = @keeps.last + keep_start = keep[0] + string = @scanner.string + keep_data = string.byteslice(keep_start, @scanner.pos - keep_start) + if keep_data + keep_buffer = keep[1] + if keep_buffer + keep_buffer << keep_data + else + keep[1] = keep_data.dup + end + end + keep[0] = 0 + end input = @inputs.first case input @@ -746,26 +729,28 @@ class CSV sample[0, 128].index(@quote_character) end - class UnoptimizedStringIO # :nodoc: - def initialize(string) - @io = StringIO.new(string, "rb:#{string.encoding}") - end + SCANNER_TEST = (ENV["CSV_PARSER_SCANNER_TEST"] == "yes") + if SCANNER_TEST + class UnoptimizedStringIO + def initialize(string) + @io = StringIO.new(string, "rb:#{string.encoding}") + end - def gets(*args) - @io.gets(*args) - end + def gets(*args) + @io.gets(*args) + end - def each_line(*args, &block) - @io.each_line(*args, &block) - end + def each_line(*args, &block) + @io.each_line(*args, &block) + end - def eof? - @io.eof? + def eof? + @io.eof? + end end - end - SCANNER_TEST = (ENV["CSV_PARSER_SCANNER_TEST"] == "yes") - if SCANNER_TEST + SCANNER_TEST_CHUNK_SIZE = + Integer((ENV["CSV_PARSER_SCANNER_TEST_CHUNK_SIZE"] || "1"), 10) def build_scanner inputs = @samples.collect do |sample| UnoptimizedStringIO.new(sample) @@ -775,12 +760,10 @@ class CSV else inputs << @input end - chunk_size = - Integer((ENV["CSV_PARSER_SCANNER_TEST_CHUNK_SIZE"] || "1"), 10) InputsScanner.new(inputs, @encoding, @row_separator, - chunk_size: chunk_size) + chunk_size: SCANNER_TEST_CHUNK_SIZE) end else def build_scanner |