diff --git a/src/pugixml.cpp b/src/pugixml.cpp index 4340c2d..4592ae7 100644 --- a/src/pugixml.cpp +++ b/src/pugixml.cpp @@ -1861,7 +1861,7 @@ PUGI__NS_BEGIN enum chartypex_t { ctx_special_pcdata = 1, // Any symbol >= 0 and < 32 (except \t, \r, \n), &, <, > - ctx_special_attr = 2, // Any symbol >= 0 and < 32 (except \t), &, <, >, " + ctx_special_attr = 2, // Any symbol >= 0 and < 32, &, <, >, " ctx_start_symbol = 4, // Any symbol > 127, a-z, A-Z, _ ctx_digit = 8, // 0-9 ctx_symbol = 16 // Any symbol > 127, a-z, A-Z, 0-9, _, -, . @@ -1869,7 +1869,7 @@ PUGI__NS_BEGIN static const unsigned char chartypex_table[256] = { - 3, 3, 3, 3, 3, 3, 3, 3, 3, 0, 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 0, 0, 2, 0, 0, 0, 3, 0, 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 diff --git a/tests/test_write.cpp b/tests/test_write.cpp index 95cc566..57fa7fb 100644 --- a/tests/test_write.cpp +++ b/tests/test_write.cpp @@ -193,7 +193,21 @@ TEST_XML(write_escape, "text") 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")); - CHECK_NODE(doc, STR("<>'\"&\r\n\t")); + CHECK_NODE(doc, STR("<>'\"&\r\n\t")); +} + +TEST_XML(write_escape_roundtrip, "text") +{ + 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")); + + std::string contents = write_narrow(doc, format_raw, encoding_utf8); + + CHECK(doc.load_buffer(contents.c_str(), contents.size())); + + // 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) + CHECK_NODE(doc, STR("<>'\"&\n\t")); } TEST_XML(write_escape_unicode, "")