External/pugixml
8
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Add support for using single quotes to enclose attribute values

Browse Source 
This change adds format_attribute_single_quote flag that uses single quotes (`'`) instead of double quotes (`"`) for formatting attribute values.

Internal quotation marks are escaped using `"` and `'`.
This commit is contained in:
m-naumann 2019-06-18 19:51:10 -07:00 committed by Arseny Kapoulkine
parent fcb7c8d3e5
commit 5a867cb1e3
5 changed files with 21 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
docs/manual.adoc
View File

@ -1706,6 +1706,8 @@ These flags control the resulting tree contents:
* [[format_skip_control_chars]]`format_skip_control_chars` enables skipping characters belonging to range [0; 32) instead of "&#xNN;" encoding. This flag is *off* by default. * [[format_skip_control_chars]]`format_skip_control_chars` enables skipping characters belonging to range [0; 32) instead of "&#xNN;" encoding. This flag is *off* by default.
* [[format_attribute_single_quote]]`format_attribute_single_quote` enables using single quotes `'` instead of double quotes `"` for enclosing attribute values. This flag is *off* by default.
These flags control the additional output information: These flags control the additional output information:
* [[format_no_declaration]]`format_no_declaration` disables default node declaration output. By default, if the document is saved via `save` or `save_file` function, and it does not have any document declaration, a default declaration is output before the document contents. Enabling this flag disables this declaration. This flag has no effect in `xml_node::print` functions: they never output the default declaration. This flag is *off* by default. * [[format_no_declaration]]`format_no_declaration` disables default node declaration output. By default, if the document is saved via `save` or `save_file` function, and it does not have any document declaration, a default declaration is output before the document contents. Enabling this flag disables this declaration. This flag has no effect in `xml_node::print` functions: they never output the default declaration. This flag is *off* by default.

3
docs/manual.html
View File

@ -3327,6 +3327,9 @@ You should use the usual bitwise arithmetics to manipulate the bitmask: to enabl
<li> <li>
<p><a id="format_skip_control_chars"></a><code>format_skip_control_chars</code> enables skipping characters belonging to range [0; 32) instead of "&amp;#xNN;" encoding. This flag is <strong>off</strong> by default.</p> <p><a id="format_skip_control_chars"></a><code>format_skip_control_chars</code> enables skipping characters belonging to range [0; 32) instead of "&amp;#xNN;" encoding. This flag is <strong>off</strong> by default.</p>
</li> </li>
<li>
<p><a id="format_attribute_single_quote"></a><code>format_attribute_single_quote</code> enables using single quotes <code>'</code> instead of double quotes <code>"</code> for enclosing attribute values. This flag is <strong>off</strong> by default.</p>
</li>
</ul> </ul>
</div> </div>
<div class="paragraph"> <div class="paragraph">

13
src/pugixml.cpp
View File

@ -1861,7 +1861,7 @@ PUGI__NS_BEGIN
enum chartypex_t enum chartypex_t
{ {
ctx_special_pcdata = 1, // Any symbol >= 0 and < 32 (except \t, \r, \n), &, <, > ctx_special_pcdata = 1, // Any symbol >= 0 and < 32 (except \t, \r, \n), &, <, >
ctx_special_attr = 2, // Any symbol >= 0 and < 32, &, <, >, " ctx_special_attr = 2, // Any symbol >= 0 and < 32, &, <, >, ", '
ctx_start_symbol = 4, // Any symbol > 127, a-z, A-Z, _ ctx_start_symbol = 4, // Any symbol > 127, a-z, A-Z, _
ctx_digit = 8, // 0-9 ctx_digit = 8, // 0-9
ctx_symbol = 16 // Any symbol > 127, a-z, A-Z, 0-9, _, -, . ctx_symbol = 16 // Any symbol > 127, a-z, A-Z, 0-9, _, -, .
@ -1871,7 +1871,7 @@ PUGI__NS_BEGIN
{ {
3, 3, 3, 3, 3, 3, 3, 3, 3, 2, 2, 3, 3, 2, 3, 3, // 0-15 3, 3, 3, 3, 3, 3, 3, 3, 3, 2, 2, 3, 3, 2, 3, 3, // 0-15
3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, // 16-31 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, // 16-31
0, 0, 2, 0, 0, 0, 3, 0, 0, 0, 0, 0, 0, 16, 16, 0, // 32-47 0, 0, 2, 0, 0, 0, 3, 2, 0, 0, 0, 0, 0, 16, 16, 0, // 32-47
24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 0, 0, 3, 0, 3, 0, // 48-63 24, 24, 24, 24, 24, 24, 24, 24, 24, 24, 0, 0, 3, 0, 3, 0, // 48-63
0, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, // 64-79 0, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, 20, // 64-79
@ -3933,6 +3933,10 @@ PUGI__NS_BEGIN
writer.write('&', 'q', 'u', 'o', 't', ';'); writer.write('&', 'q', 'u', 'o', 't', ';');
++s; ++s;
break; break;
case '\'':
writer.write('&', 'a', 'p', 'o', 's', ';');
++s;
break;
default: // s is not a usual symbol default: // s is not a usual symbol
{ {
unsigned int ch = static_cast<unsigned int>(*s++); unsigned int ch = static_cast<unsigned int>(*s++);
@ -4064,6 +4068,7 @@ PUGI__NS_BEGIN
PUGI__FN void node_output_attributes(xml_buffered_writer& writer, xml_node_struct* node, const char_t* indent, size_t indent_length, unsigned int flags, unsigned int depth) PUGI__FN void node_output_attributes(xml_buffered_writer& writer, xml_node_struct* node, const char_t* indent, size_t indent_length, unsigned int flags, unsigned int depth)
{ {
const char_t* default_name = PUGIXML_TEXT(":anonymous"); const char_t* default_name = PUGIXML_TEXT(":anonymous");
const char_t enquotation_char = (flags & format_attribute_single_quote) ? '\'' : '"';
for (xml_attribute_struct* a = node->first_attribute; a; a = a->next_attribute) for (xml_attribute_struct* a = node->first_attribute; a; a = a->next_attribute)
{ {
@ -4079,12 +4084,12 @@ PUGI__NS_BEGIN
} }
writer.write_string(a->name ? a->name + 0 : default_name); writer.write_string(a->name ? a->name + 0 : default_name);
writer.write('=', '"'); writer.write('=', enquotation_char);
if (a->value) if (a->value)
text_output(writer, a->value, ctx_special_attr, flags); text_output(writer, a->value, ctx_special_attr, flags);
writer.write('"'); writer.write(enquotation_char);
} }
} }

3
src/pugixml.hpp
View File

@ -255,6 +255,9 @@ namespace pugi
// Skip characters belonging to range [0; 32) instead of "&#xNN;" encoding. This flag is off by default. // Skip characters belonging to range [0; 32) instead of "&#xNN;" encoding. This flag is off by default.
const unsigned int format_skip_control_chars = 0x100; const unsigned int format_skip_control_chars = 0x100;
// Use single quotes ' instead of double quotes " for enclosing attribute values. This flag is off by default.
const unsigned int format_attribute_single_quote = 0x200;
// The default set of formatting flags. // The default set of formatting flags.
// Nodes are indented depending on their depth in DOM tree, a default declaration is output if document has none. // Nodes are indented depending on their depth in DOM tree, a default declaration is output if document has none.
const unsigned int format_default = format_indent; const unsigned int format_default = format_indent;

6
tests/test_write.cpp
View File

@ -193,7 +193,7 @@ TEST_XML(write_escape, "<node attr=''>text</node>")
doc.child(STR("node")).attribute(STR("attr")) = STR("<>'\"&\x04\r\n\t"); doc.child(STR("node")).attribute(STR("attr")) = STR("<>'\"&\x04\r\n\t");
doc.child(STR("node")).first_child().set_value(STR("<>'\"&\x04\r\n\t")); doc.child(STR("node")).first_child().set_value(STR("<>'\"&\x04\r\n\t"));
CHECK_NODE(doc, STR("<node attr=\"&lt;&gt;'&quot;&amp;&#04;&#13;&#10;&#09;\">&lt;&gt;'\"&amp;&#04;\r\n\t</node>")); CHECK_NODE(doc, STR("<node attr=\"&lt;&gt;&apos;&quot;&amp;&#04;&#13;&#10;&#09;\">&lt;&gt;'\"&amp;&#04;\r\n\t</node>"));
} }
TEST_XML(write_escape_roundtrip, "<node attr=''>text</node>") TEST_XML(write_escape_roundtrip, "<node attr=''>text</node>")
@ -207,7 +207,7 @@ TEST_XML(write_escape_roundtrip, "<node attr=''>text</node>")
// Note: this string is almost identical to the string from write_escape with the exception of \r // Note: this string is almost identical to the string from write_escape with the exception of \r
// \r in PCDATA doesn't roundtrip because it has to go through newline conversion (which could be disabled, but is active by default) // \r in PCDATA doesn't roundtrip because it has to go through newline conversion (which could be disabled, but is active by default)
CHECK_NODE(doc, STR("<node attr=\"&lt;&gt;'&quot;&amp;&#04;&#13;&#10;&#09;\">&lt;&gt;'\"&amp;&#04;\n\t</node>")); CHECK_NODE(doc, STR("<node attr=\"&lt;&gt;&apos;&quot;&amp;&#04;&#13;&#10;&#09;\">&lt;&gt;'\"&amp;&#04;\n\t</node>"));
} }
TEST_XML(write_escape_unicode, "<node attr='&#x3c00;'/>") TEST_XML(write_escape_unicode, "<node attr='&#x3c00;'/>")
@ -632,7 +632,7 @@ TEST_XML(write_no_empty_element_tags, "<node><child1/><child2>text</child2><chil
TEST_XML_FLAGS(write_roundtrip, "<node><child1 attr1='value1' attr2='value2'/><child2 attr='value'>pre<![CDATA[data]]>mid&lt;text&amp;escape<!--comment--><test/>post<?pi value?>fin</child2><child3/></node>", parse_full) TEST_XML_FLAGS(write_roundtrip, "<node><child1 attr1='value1' attr2='value2'/><child2 attr='value'>pre<![CDATA[data]]>mid&lt;text&amp;escape<!--comment--><test/>post<?pi value?>fin</child2><child3/></node>", parse_full)
{ {
const unsigned int flagset[] = { format_indent, format_raw, format_no_declaration, format_indent_attributes, format_no_empty_element_tags }; const unsigned int flagset[] = { format_indent, format_raw, format_no_declaration, format_indent_attributes, format_no_empty_element_tags, format_attribute_single_quote };
size_t flagcount = sizeof(flagset) / sizeof(flagset[0]); size_t flagcount = sizeof(flagset) / sizeof(flagset[0]);
for (size_t i = 0; i < (size_t(1) << flagcount); ++i) for (size_t i = 0; i < (size_t(1) << flagcount); ++i)