OLD | NEW |
(Empty) | |
| 1 #include "Filter.h" |
| 2 #include "CommentFilter.h" |
| 3 #include "InvalidFilter.h" |
| 4 #include "RegExpFilter.h" |
| 5 #include "BlockingFilter.h" |
| 6 #include "WhitelistFilter.h" |
| 7 #include "ElemHideBase.h" |
| 8 #include "ElemHideFilter.h" |
| 9 #include "ElemHideException.h" |
| 10 #include "CSSPropertyFilter.h" |
| 11 #include "StringMap.h" |
| 12 |
| 13 namespace |
| 14 { |
| 15 StringMap<Filter*> knownFilters(8192); |
| 16 |
| 17 void NormalizeWhitespace(DependentString& text) |
| 18 { |
| 19 String::size_type start = 0; |
| 20 String::size_type end = text.length(); |
| 21 |
| 22 // Remove leading spaces and special characters like line breaks |
| 23 for (; start < end; start++) |
| 24 if (text[start] > ' ') |
| 25 break; |
| 26 |
| 27 // Now look for invalid characters inside the string |
| 28 String::size_type pos; |
| 29 for (pos = start; pos < end; pos++) |
| 30 if (text[pos] < ' ') |
| 31 break; |
| 32 |
| 33 if (pos < end) |
| 34 { |
| 35 // Found invalid characters, copy all the valid characters while skipping |
| 36 // the invalid ones. |
| 37 String::size_type delta = 1; |
| 38 for (pos = pos + 1; pos < end; pos++) |
| 39 { |
| 40 if (text[pos] < ' ') |
| 41 delta++; |
| 42 else |
| 43 text[pos - delta] = text[pos]; |
| 44 } |
| 45 end -= delta; |
| 46 } |
| 47 |
| 48 // Remove trailing spaces |
| 49 for (; end > 0; end--) |
| 50 if (text[end - 1] != ' ') |
| 51 break; |
| 52 |
| 53 // Set new string boundaries |
| 54 text.reset(text, start, end - start); |
| 55 } |
| 56 } |
| 57 |
| 58 Filter::Filter(Type type, const String& text) |
| 59 : mType(type), mText(text) |
| 60 { |
| 61 annotate_address(this, "Filter"); |
| 62 } |
| 63 |
| 64 Filter::~Filter() |
| 65 { |
| 66 knownFilters.erase(mText); |
| 67 } |
| 68 |
| 69 OwnedString Filter::Serialize() const |
| 70 { |
| 71 OwnedString result(u"[Filter]\ntext="_str); |
| 72 result.append(mText); |
| 73 result.append(u'\n'); |
| 74 return result; |
| 75 } |
| 76 |
| 77 Filter* Filter::FromText(DependentString& text) |
| 78 { |
| 79 NormalizeWhitespace(text); |
| 80 if (text.empty()) |
| 81 return nullptr; |
| 82 |
| 83 // Parsing also normalizes the filter text, so it has to be done before the |
| 84 // lookup in knownFilters. |
| 85 union |
| 86 { |
| 87 RegExpFilterData regexp; |
| 88 ElemHideData elemhide; |
| 89 } data; |
| 90 DependentString error; |
| 91 |
| 92 Filter::Type type = CommentFilter::Parse(text); |
| 93 if (type == Filter::Type::UNKNOWN) |
| 94 type = ElemHideBase::Parse(text, data.elemhide); |
| 95 if (type == Filter::Type::UNKNOWN) |
| 96 type = RegExpFilter::Parse(text, error, data.regexp); |
| 97 |
| 98 auto knownFilter = knownFilters.find(text); |
| 99 if (knownFilter) |
| 100 { |
| 101 knownFilter->second->AddRef(); |
| 102 return knownFilter->second; |
| 103 } |
| 104 |
| 105 FilterPtr filter; |
| 106 switch (type) |
| 107 { |
| 108 case Filter::Type::COMMENT: |
| 109 filter = new CommentFilter(text); |
| 110 break; |
| 111 case Filter::Type::INVALID: |
| 112 filter = new InvalidFilter(text, error); |
| 113 break; |
| 114 case Filter::Type::BLOCKING: |
| 115 filter = new BlockingFilter(text, data.regexp); |
| 116 break; |
| 117 case Filter::Type::WHITELIST: |
| 118 filter = new WhitelistFilter(text, data.regexp); |
| 119 break; |
| 120 case Filter::Type::ELEMHIDE: |
| 121 filter = new ElemHideFilter(text, data.elemhide); |
| 122 break; |
| 123 case Filter::Type::ELEMHIDEEXCEPTION: |
| 124 filter = new ElemHideException(text, data.elemhide); |
| 125 break; |
| 126 case Filter::Type::CSSPROPERTY: |
| 127 filter = new CSSPropertyFilter(text, data.elemhide); |
| 128 if (static_cast<CSSPropertyFilter*>(filter.get())->IsGeneric()) |
| 129 filter = new InvalidFilter(text, u"filter_cssproperty_nodomain"_str); |
| 130 break; |
| 131 default: |
| 132 // This should never happen but just in case |
| 133 return nullptr; |
| 134 } |
| 135 |
| 136 // This is a hack: we looked up the entry using text but create it using |
| 137 // filter->mText. This works because both are equal at this point. However, |
| 138 // text refers to a temporary buffer which will go away. |
| 139 enter_context("Adding to known filters"); |
| 140 knownFilter.assign(filter->mText, filter.get()); |
| 141 exit_context(); |
| 142 |
| 143 return filter.release(); |
| 144 } |
OLD | NEW |