nokogumbo 0.5 → 0.5.1
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- data/work/Makefile +213 -0
- data/work/attribute.c +44 -0
- data/work/attribute.h +37 -0
- data/work/attribute.o +0 -0
- data/work/char_ref.c +2561 -0
- data/work/char_ref.h +61 -0
- data/work/char_ref.o +0 -0
- data/work/error.c +258 -0
- data/work/error.h +225 -0
- data/work/error.o +0 -0
- data/work/gumbo.h +800 -0
- data/work/insertion_mode.h +54 -0
- data/work/mkmf.log +41 -0
- data/work/nokogumbo.c +97 -0
- data/work/nokogumbo.o +0 -0
- data/work/nokogumboc.so +0 -0
- data/work/parser.c +3893 -0
- data/work/parser.h +57 -0
- data/work/parser.o +0 -0
- data/work/string_buffer.c +106 -0
- data/work/string_buffer.h +82 -0
- data/work/string_buffer.o +0 -0
- data/work/string_piece.c +49 -0
- data/work/string_piece.h +39 -0
- data/work/string_piece.o +0 -0
- data/work/tag.c +222 -0
- data/work/tag.o +0 -0
- data/work/token_type.h +40 -0
- data/work/tokenizer.c +2978 -0
- data/work/tokenizer.h +123 -0
- data/work/tokenizer.o +0 -0
- data/work/tokenizer_states.h +103 -0
- data/work/utf8.c +268 -0
- data/work/utf8.h +127 -0
- data/work/utf8.o +0 -0
- data/work/util.c +58 -0
- data/work/util.h +57 -0
- data/work/util.o +0 -0
- data/work/vector.c +121 -0
- data/work/vector.h +66 -0
- data/work/vector.o +0 -0
- metadata +42 -2
- data/Rakefile +0 -68
data/work/Makefile
ADDED
@@ -0,0 +1,213 @@
|
|
1
|
+
|
2
|
+
SHELL = /bin/sh
|
3
|
+
|
4
|
+
# V=0 quiet, V=1 verbose. other values don't work.
|
5
|
+
V = 0
|
6
|
+
Q1 = $(V:1=)
|
7
|
+
Q = $(Q1:0=@)
|
8
|
+
n=$(NULLCMD)
|
9
|
+
ECHO1 = $(V:1=@$n)
|
10
|
+
ECHO = $(ECHO1:0=@echo)
|
11
|
+
|
12
|
+
#### Start of system configuration section. ####
|
13
|
+
|
14
|
+
srcdir = .
|
15
|
+
topdir = /usr/include/ruby-1.9.1
|
16
|
+
hdrdir = /usr/include/ruby-1.9.1
|
17
|
+
arch_hdrdir = /usr/include/ruby-1.9.1/$(arch)
|
18
|
+
VPATH = $(srcdir):$(arch_hdrdir)/ruby:$(hdrdir)/ruby
|
19
|
+
prefix = $(DESTDIR)/usr
|
20
|
+
rubylibprefix = $(libdir)/$(RUBY_BASE_NAME)
|
21
|
+
exec_prefix = $(prefix)
|
22
|
+
vendorhdrdir = $(rubyhdrdir)/vendor_ruby
|
23
|
+
sitehdrdir = $(rubyhdrdir)/site_ruby
|
24
|
+
rubyhdrdir = $(includedir)/$(RUBY_BASE_NAME)-$(ruby_version)
|
25
|
+
vendordir = $(DESTDIR)/usr/lib/ruby/vendor_ruby
|
26
|
+
sitedir = $(DESTDIR)/usr/local/lib/site_ruby
|
27
|
+
ridir = $(datarootdir)/$(RI_BASE_NAME)
|
28
|
+
mandir = $(prefix)/share/man
|
29
|
+
localedir = $(datarootdir)/locale
|
30
|
+
libdir = $(exec_prefix)/lib
|
31
|
+
psdir = $(docdir)
|
32
|
+
pdfdir = $(docdir)
|
33
|
+
dvidir = $(docdir)
|
34
|
+
htmldir = $(docdir)
|
35
|
+
infodir = $(prefix)/share/info
|
36
|
+
docdir = $(datarootdir)/doc/$(PACKAGE)
|
37
|
+
oldincludedir = $(DESTDIR)/usr/include
|
38
|
+
includedir = $(prefix)/include
|
39
|
+
localstatedir = $(DESTDIR)/var
|
40
|
+
sharedstatedir = $(prefix)/com
|
41
|
+
sysconfdir = $(DESTDIR)/etc
|
42
|
+
datadir = $(datarootdir)
|
43
|
+
datarootdir = $(prefix)/share
|
44
|
+
libexecdir = $(prefix)/lib/ruby1.9.1
|
45
|
+
sbindir = $(exec_prefix)/sbin
|
46
|
+
bindir = $(exec_prefix)/bin
|
47
|
+
rubylibdir = $(rubylibprefix)/$(ruby_version)
|
48
|
+
archdir = $(rubylibdir)/$(arch)
|
49
|
+
sitelibdir = $(sitedir)/$(ruby_version)
|
50
|
+
sitearchdir = $(sitelibdir)/$(sitearch)
|
51
|
+
vendorlibdir = $(vendordir)/$(ruby_version)
|
52
|
+
vendorarchdir = $(vendorlibdir)/$(sitearch)
|
53
|
+
|
54
|
+
NULLCMD = :
|
55
|
+
|
56
|
+
CC = gcc
|
57
|
+
CXX = g++
|
58
|
+
LIBRUBY = $(LIBRUBY_SO)
|
59
|
+
LIBRUBY_A = lib$(RUBY_SO_NAME)-static.a
|
60
|
+
LIBRUBYARG_SHARED = -l$(RUBY_SO_NAME)
|
61
|
+
LIBRUBYARG_STATIC = -l$(RUBY_SO_NAME)-static
|
62
|
+
OUTFLAG = -o
|
63
|
+
COUTFLAG = -o
|
64
|
+
|
65
|
+
RUBY_EXTCONF_H =
|
66
|
+
cflags = $(optflags) $(debugflags) $(warnflags)
|
67
|
+
optflags = -O3
|
68
|
+
debugflags = -ggdb
|
69
|
+
warnflags = -Wextra -Wno-unused-parameter -Wno-parentheses -Wno-long-long -Wno-missing-field-initializers -Wpointer-arith -Wwrite-strings -Wdeclaration-after-statement -Wimplicit-function-declaration
|
70
|
+
CFLAGS = -fPIC -std=c99 -I/usr/include/libxml2 $(ARCH_FLAG)
|
71
|
+
INCFLAGS = -I. -I$(arch_hdrdir) -I$(hdrdir)/ruby/backward -I$(hdrdir) -I$(srcdir) -I/var/lib/gems/1.9.1/gems/nokogiri-1.5.5/ext/nokogiri
|
72
|
+
DEFS =
|
73
|
+
CPPFLAGS = -D_FORTIFY_SOURCE=2 $(DEFS) $(cppflags)
|
74
|
+
CXXFLAGS = $(CFLAGS) -g -O2 -fstack-protector --param=ssp-buffer-size=4 -Wformat -Werror=format-security -Wall -fno-strict-aliasing
|
75
|
+
ldflags = -L. -Wl,-Bsymbolic-functions -Wl,-z,relro -rdynamic -Wl,-export-dynamic
|
76
|
+
dldflags =
|
77
|
+
ARCH_FLAG =
|
78
|
+
DLDFLAGS = $(ldflags) $(dldflags) $(ARCH_FLAG)
|
79
|
+
LDSHARED = $(CC) -shared
|
80
|
+
LDSHAREDXX = $(CXX) -shared
|
81
|
+
AR = ar
|
82
|
+
EXEEXT =
|
83
|
+
|
84
|
+
RUBY_BASE_NAME = ruby
|
85
|
+
RUBY_INSTALL_NAME = ruby1.9.1
|
86
|
+
RUBY_SO_NAME = ruby-1.9.1
|
87
|
+
arch = x86_64-linux
|
88
|
+
sitearch = $(arch)
|
89
|
+
ruby_version = 1.9.1
|
90
|
+
ruby = /usr/bin/ruby1.9.1
|
91
|
+
RUBY = $(ruby)
|
92
|
+
RM = rm -f
|
93
|
+
RM_RF = $(RUBY) -run -e rm -- -rf
|
94
|
+
RMDIRS = rmdir --ignore-fail-on-non-empty -p
|
95
|
+
MAKEDIRS = /bin/mkdir -p
|
96
|
+
INSTALL = /usr/bin/install -c
|
97
|
+
INSTALL_PROG = $(INSTALL) -m 0755
|
98
|
+
INSTALL_DATA = $(INSTALL) -m 644
|
99
|
+
COPY = cp
|
100
|
+
|
101
|
+
#### End of system configuration section. ####
|
102
|
+
|
103
|
+
preload =
|
104
|
+
|
105
|
+
libpath = . $(libdir)
|
106
|
+
LIBPATH = -L. -L$(libdir)
|
107
|
+
DEFFILE =
|
108
|
+
|
109
|
+
CLEANFILES = mkmf.log
|
110
|
+
DISTCLEANFILES =
|
111
|
+
DISTCLEANDIRS =
|
112
|
+
|
113
|
+
extout =
|
114
|
+
extout_prefix =
|
115
|
+
target_prefix =
|
116
|
+
LOCAL_LIBS =
|
117
|
+
LIBS = $(LIBRUBYARG_SHARED) -lxml2 -lpthread -lrt -ldl -lcrypt -lm -lc
|
118
|
+
SRCS = nokogumbo.c vector.c string_buffer.c error.c attribute.c char_ref.c string_piece.c tag.c util.c parser.c utf8.c tokenizer.c
|
119
|
+
OBJS = nokogumbo.o vector.o string_buffer.o error.o attribute.o char_ref.o string_piece.o tag.o util.o parser.o utf8.o tokenizer.o
|
120
|
+
TARGET = nokogumboc
|
121
|
+
DLLIB = $(TARGET).so
|
122
|
+
EXTSTATIC =
|
123
|
+
STATIC_LIB =
|
124
|
+
|
125
|
+
BINDIR = $(bindir)
|
126
|
+
RUBYCOMMONDIR = $(sitedir)$(target_prefix)
|
127
|
+
RUBYLIBDIR = $(sitelibdir)$(target_prefix)
|
128
|
+
RUBYARCHDIR = $(sitearchdir)$(target_prefix)
|
129
|
+
HDRDIR = $(rubyhdrdir)/ruby$(target_prefix)
|
130
|
+
ARCHHDRDIR = $(rubyhdrdir)/$(arch)/ruby$(target_prefix)
|
131
|
+
|
132
|
+
TARGET_SO = $(DLLIB)
|
133
|
+
CLEANLIBS = $(TARGET).so
|
134
|
+
CLEANOBJS = *.o *.bak
|
135
|
+
|
136
|
+
all: $(DLLIB)
|
137
|
+
static: $(STATIC_LIB)
|
138
|
+
.PHONY: all install static install-so install-rb
|
139
|
+
.PHONY: clean clean-so clean-rb
|
140
|
+
|
141
|
+
clean-rb-default::
|
142
|
+
clean-rb::
|
143
|
+
clean-so::
|
144
|
+
clean: clean-so clean-rb-default clean-rb
|
145
|
+
@-$(RM) $(CLEANLIBS) $(CLEANOBJS) $(CLEANFILES)
|
146
|
+
|
147
|
+
distclean-rb-default::
|
148
|
+
distclean-rb::
|
149
|
+
distclean-so::
|
150
|
+
distclean: clean distclean-so distclean-rb-default distclean-rb
|
151
|
+
@-$(RM) Makefile $(RUBY_EXTCONF_H) conftest.* mkmf.log
|
152
|
+
@-$(RM) core ruby$(EXEEXT) *~ $(DISTCLEANFILES)
|
153
|
+
@-$(RMDIRS) $(DISTCLEANDIRS) 2> /dev/null || true
|
154
|
+
|
155
|
+
realclean: distclean
|
156
|
+
install: install-so install-rb
|
157
|
+
|
158
|
+
install-so: $(RUBYARCHDIR)
|
159
|
+
install-so: $(RUBYARCHDIR)/$(DLLIB)
|
160
|
+
$(RUBYARCHDIR)/$(DLLIB): $(DLLIB)
|
161
|
+
@-$(MAKEDIRS) $(@D)
|
162
|
+
$(INSTALL_PROG) $(DLLIB) $(@D)
|
163
|
+
install-rb: pre-install-rb install-rb-default
|
164
|
+
install-rb-default: pre-install-rb-default
|
165
|
+
pre-install-rb: Makefile
|
166
|
+
pre-install-rb-default: Makefile
|
167
|
+
pre-install-rb-default:
|
168
|
+
$(ECHO) installing default nokogumboc libraries
|
169
|
+
$(RUBYARCHDIR):
|
170
|
+
$(Q) $(MAKEDIRS) $@
|
171
|
+
|
172
|
+
site-install: site-install-so site-install-rb
|
173
|
+
site-install-so: install-so
|
174
|
+
site-install-rb: install-rb
|
175
|
+
|
176
|
+
.SUFFIXES: .c .m .cc .mm .cxx .cpp .C .o
|
177
|
+
|
178
|
+
.cc.o:
|
179
|
+
$(ECHO) compiling $(<)
|
180
|
+
$(Q) $(CXX) $(INCFLAGS) $(CPPFLAGS) $(CXXFLAGS) $(COUTFLAG)$@ -c $<
|
181
|
+
|
182
|
+
.mm.o:
|
183
|
+
$(ECHO) compiling $(<)
|
184
|
+
$(Q) $(CXX) $(INCFLAGS) $(CPPFLAGS) $(CXXFLAGS) $(COUTFLAG)$@ -c $<
|
185
|
+
|
186
|
+
.cxx.o:
|
187
|
+
$(ECHO) compiling $(<)
|
188
|
+
$(Q) $(CXX) $(INCFLAGS) $(CPPFLAGS) $(CXXFLAGS) $(COUTFLAG)$@ -c $<
|
189
|
+
|
190
|
+
.cpp.o:
|
191
|
+
$(ECHO) compiling $(<)
|
192
|
+
$(Q) $(CXX) $(INCFLAGS) $(CPPFLAGS) $(CXXFLAGS) $(COUTFLAG)$@ -c $<
|
193
|
+
|
194
|
+
.C.o:
|
195
|
+
$(ECHO) compiling $(<)
|
196
|
+
$(Q) $(CXX) $(INCFLAGS) $(CPPFLAGS) $(CXXFLAGS) $(COUTFLAG)$@ -c $<
|
197
|
+
|
198
|
+
.c.o:
|
199
|
+
$(ECHO) compiling $(<)
|
200
|
+
$(Q) $(CC) $(INCFLAGS) $(CPPFLAGS) $(CFLAGS) $(COUTFLAG)$@ -c $<
|
201
|
+
|
202
|
+
.m.o:
|
203
|
+
$(ECHO) compiling $(<)
|
204
|
+
$(Q) $(CC) $(INCFLAGS) $(CPPFLAGS) $(CFLAGS) $(COUTFLAG)$@ -c $<
|
205
|
+
|
206
|
+
$(DLLIB): $(OBJS) Makefile
|
207
|
+
$(ECHO) linking shared-object $(DLLIB)
|
208
|
+
@-$(RM) $(@)
|
209
|
+
$(Q) $(LDSHARED) -o $@ $(OBJS) $(LIBPATH) $(DLDFLAGS) $(LOCAL_LIBS) $(LIBS)
|
210
|
+
|
211
|
+
|
212
|
+
|
213
|
+
$(OBJS): $(hdrdir)/ruby.h $(hdrdir)/ruby/defines.h $(arch_hdrdir)/ruby/config.h
|
data/work/attribute.c
ADDED
@@ -0,0 +1,44 @@
|
|
1
|
+
// Copyright 2010 Google Inc. All Rights Reserved.
|
2
|
+
//
|
3
|
+
// Licensed under the Apache License, Version 2.0 (the "License");
|
4
|
+
// you may not use this file except in compliance with the License.
|
5
|
+
// You may obtain a copy of the License at
|
6
|
+
//
|
7
|
+
// http://www.apache.org/licenses/LICENSE-2.0
|
8
|
+
//
|
9
|
+
// Unless required by applicable law or agreed to in writing, software
|
10
|
+
// distributed under the License is distributed on an "AS IS" BASIS,
|
11
|
+
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
12
|
+
// See the License for the specific language governing permissions and
|
13
|
+
// limitations under the License.
|
14
|
+
//
|
15
|
+
// Author: jdtang@google.com (Jonathan Tang)
|
16
|
+
|
17
|
+
#include "attribute.h"
|
18
|
+
|
19
|
+
#include <assert.h>
|
20
|
+
#include <stdlib.h>
|
21
|
+
#include <string.h>
|
22
|
+
#include <strings.h>
|
23
|
+
|
24
|
+
#include "util.h"
|
25
|
+
|
26
|
+
struct _GumboParser;
|
27
|
+
|
28
|
+
GumboAttribute* gumbo_get_attribute(
|
29
|
+
const struct _GumboVector* attributes, const char* name) {
|
30
|
+
for (int i = 0; i < attributes->length; ++i) {
|
31
|
+
GumboAttribute* attr = attributes->data[i];
|
32
|
+
if (!strcasecmp(attr->name, name)) {
|
33
|
+
return attr;
|
34
|
+
}
|
35
|
+
}
|
36
|
+
return NULL;
|
37
|
+
}
|
38
|
+
|
39
|
+
void gumbo_destroy_attribute(
|
40
|
+
struct _GumboParser* parser, GumboAttribute* attribute) {
|
41
|
+
gumbo_parser_deallocate(parser, (void*) attribute->name);
|
42
|
+
gumbo_parser_deallocate(parser, (void*) attribute->value);
|
43
|
+
gumbo_parser_deallocate(parser, (void*) attribute);
|
44
|
+
}
|
data/work/attribute.h
ADDED
@@ -0,0 +1,37 @@
|
|
1
|
+
// Copyright 2010 Google Inc. All Rights Reserved.
|
2
|
+
//
|
3
|
+
// Licensed under the Apache License, Version 2.0 (the "License");
|
4
|
+
// you may not use this file except in compliance with the License.
|
5
|
+
// You may obtain a copy of the License at
|
6
|
+
//
|
7
|
+
// http://www.apache.org/licenses/LICENSE-2.0
|
8
|
+
//
|
9
|
+
// Unless required by applicable law or agreed to in writing, software
|
10
|
+
// distributed under the License is distributed on an "AS IS" BASIS,
|
11
|
+
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
12
|
+
// See the License for the specific language governing permissions and
|
13
|
+
// limitations under the License.
|
14
|
+
//
|
15
|
+
// Author: jdtang@google.com (Jonathan Tang)
|
16
|
+
|
17
|
+
#ifndef GUMBO_ATTRIBUTE_H_
|
18
|
+
#define GUMBO_ATTRIBUTE_H_
|
19
|
+
|
20
|
+
#include "gumbo.h"
|
21
|
+
|
22
|
+
#ifdef __cplusplus
|
23
|
+
extern "C" {
|
24
|
+
#endif
|
25
|
+
|
26
|
+
struct _GumboParser;
|
27
|
+
|
28
|
+
// Release the memory used for an GumboAttribute, including the attribute
|
29
|
+
// itself.
|
30
|
+
void gumbo_destroy_attribute(
|
31
|
+
struct _GumboParser* parser, GumboAttribute* attribute);
|
32
|
+
|
33
|
+
#ifdef __cplusplus
|
34
|
+
}
|
35
|
+
#endif
|
36
|
+
|
37
|
+
#endif // GUMBO_ATTRIBUTE_H_
|
data/work/attribute.o
ADDED
Binary file
|