Code:
/ Dotnetfx_Vista_SP2 / Dotnetfx_Vista_SP2 / 8.0.50727.4016 / DEVDIV / depot / DevDiv / releases / whidbey / NetFxQFE / ndp / clr / src / BCL / System / Text / EUCJPEncoding.cs / 1 / EUCJPEncoding.cs
// ==++== // // Copyright (c) Microsoft Corporation. All rights reserved. // // ==--== namespace System.Text { using System.Text; using System.Globalization; // EUCJPEncoding // // EUC-JP Encoding (51932) // // EUC-JP has the following code points: // 00-7F - ASCII // 80-8D & 90-9F - Control. (Like Unicode, except for 8e and 8f) // A1-FE, A1-FE - 2 byte JIS X 0208 range. // 8E, A1-DF - 2 byte half-width Katakana // 8F, A1-FE, A1-FE - 3 byte JIX X 0212 range. WE DON'T USE JIS 0212!!! // // New thoughts: // Fixing windows 20932 code page so that all characters can be looked up there. // // Old thoughts: // Windows NLS uses a special CP20932 for EUC-JP, but it is not used by mlang. Windows // Maps the 3 byte ranges to the 2 byte CP20932 by masking the 2nd byte with & 0x7F. // MLang uses the native windows 932 code page, which is more reliable, however the code points // don't line up as nicely as the 20932 code page, however it doesn't have JIS X 0212 support. // // So what we do is: // 1. For ASCII, leave it alone // 2. For half-width Katakana, use the leading byte and convert with 20936 code page. // 3. For JIS X 0208, Use the leading & trailing bytes with 20936 code page // 4. For JIS X 0212, Remove the lead byte, & 0xFF7F, and use the CP20936 table to convert. // // Regarding Normalization: // Forms KC & KD are precluded because of things like halfwidth Katakana that has compatibility mappings // Form D is precluded because of 0x00a8, which changes to space + dierises. // // I think that IsAlwaysNormalized should probably return true for form C (but not certain) // // NOTE: We don't use JIS 0212 so we are basically a DBCS code page, we just have to modify // the 932 table we're basing this on. // using System; [Serializable()] internal class EUCJPEncoding : DBCSCodePageEncoding { // This pretends to be CP 932 as far as memory tables are concerned. public EUCJPEncoding() : base(51932, 932) { this.m_bUseMlangTypeForSerialization = true; } protected unsafe override String GetMemorySectionName() { int iUseCodePage = this.bFlagDataTable ? dataTableCodePage : CodePage; String strName = String.Format(CultureInfo.InvariantCulture, "CodePage_{0}_{1}_{2}_{3}_{4}_EUCJP", iUseCodePage, this.pCodePage->VersionMajor, this.pCodePage->VersionMinor, this.pCodePage->VersionRevision, this.pCodePage->VersionBuild); return strName; } // Clean up characters for EUC-JP code pages, etc. protected override bool CleanUpBytes(ref int bytes) { if (bytes >= 0x100) { // map extended char (0xfa40-0xfc4b) to a special range // (ported from mlang) if (bytes >= 0xfa40 && bytes <= 0xfc4b) { if ( bytes >= 0xfa40 && bytes <= 0xfa5b ) { if ( bytes <= 0xfa49 ) bytes = bytes - 0x0b51 ; else if ( bytes >= 0xfa4a && bytes <= 0xfa53 ) bytes = bytes - 0x072f6 ; else if ( bytes >= 0xfa54 && bytes <= 0xfa57 ) bytes = bytes - 0x0b5b ; else if ( bytes == 0xfa58 ) bytes = 0x878a ; else if ( bytes == 0xfa59 ) bytes = 0x8782 ; else if ( bytes == 0xfa5a ) bytes = 0x8784 ; else if ( bytes == 0xfa5b ) bytes = 0x879a ; } else if ( bytes >= 0xfa5c && bytes <= 0xfc4b ) { byte tc = unchecked((byte)bytes); if ( tc < 0x5c ) bytes = bytes - 0x0d5f; else if ( tc >= 0x80 && tc <= 0x9B ) bytes = bytes - 0x0d1d; else bytes = bytes - 0x0d1c; } } // Convert 932 code page to 20932 like code page range // (also ported from mlang) byte bLead = unchecked((byte)(bytes >> 8)); byte bTrail = unchecked((byte)bytes); bLead -= ((bLead > (byte)0x9f) ? (byte)0xb1 : (byte)0x71); bLead = (byte)((bLead << 1) + 1); if (bTrail > (byte)0x9e) { bTrail -= (byte)0x7e; bLead++; } else { if (bTrail > (byte)0x7e) bTrail--; bTrail -= (byte)0x1f; } bytes = ((int)bLead) << 8 | (int)bTrail | 0x8080; // // Don't step on our katakana special hack plane, if katakana space return false. // if (bytes >= 0x8E00 && bytes <= 0x8EFF) // return false; // Don't step out of our allocated lead byte area. // All DBCS lead and trail bytes should be >= 0xa1 and <= 0xfe if ((bytes & 0xFF00) < 0xa100 || (bytes & 0xFF00) > 0xfe00 || (bytes & 0xFF) < 0xa1 || (bytes & 0xFF) > 0xfe) return false; // WARNING: Our funky mapping allows illegal values, which we continue to use // so that we're compatible with Everett. } else { // For 51932 1/2 Katakana gets a 0x8E lead byte // Adjust 1/2 Katakana if (bytes >= 0xa1 && bytes <= 0xdf) { bytes |= 0x8E00; return true; } // 0x81-0x9f and 0xe0-0xfc CP 932 // 0x8e and 0xa1-0xfe CP 20932 (we don't use 8e though) // b0-df is 1/2 Katakana // So 81-9f & e0-fc are 932 lead bytes, a1-fe are our lead bytes // so ignore everything above 0x80 except 0xa0 and 0xff if (bytes >= 0x81 && bytes != 0xa0 && bytes != 0xff) { // We set diffent lead bytes later, so just return false return false; } } return true; } protected override unsafe void CleanUpEndBytes(char* chars) { // Need to special case CP 51932 // 0x81-0x9f and 0xe0-0xfc CP 932 // 0x8e and 0xa1-0xfe CP 20932 // 0x10 and 0x21-0x9? Us (remapping 932) // b0-df is 1/2 Katakana (trail byte) // A1-FE are DBCS code points for (int i = 0xA1; i <= 0xFE; i++) chars[i] = LEAD_BYTE_CHAR; // And 8E is lead byte for Katakana (already set) chars[0x8e] = LEAD_BYTE_CHAR; } } } // File provided for Reference Use Only by Microsoft Corporation (c) 2007. // ==++== // // Copyright (c) Microsoft Corporation. All rights reserved. // // ==--== namespace System.Text { using System.Text; using System.Globalization; // EUCJPEncoding // // EUC-JP Encoding (51932) // // EUC-JP has the following code points: // 00-7F - ASCII // 80-8D & 90-9F - Control. (Like Unicode, except for 8e and 8f) // A1-FE, A1-FE - 2 byte JIS X 0208 range. // 8E, A1-DF - 2 byte half-width Katakana // 8F, A1-FE, A1-FE - 3 byte JIX X 0212 range. WE DON'T USE JIS 0212!!! // // New thoughts: // Fixing windows 20932 code page so that all characters can be looked up there. // // Old thoughts: // Windows NLS uses a special CP20932 for EUC-JP, but it is not used by mlang. Windows // Maps the 3 byte ranges to the 2 byte CP20932 by masking the 2nd byte with & 0x7F. // MLang uses the native windows 932 code page, which is more reliable, however the code points // don't line up as nicely as the 20932 code page, however it doesn't have JIS X 0212 support. // // So what we do is: // 1. For ASCII, leave it alone // 2. For half-width Katakana, use the leading byte and convert with 20936 code page. // 3. For JIS X 0208, Use the leading & trailing bytes with 20936 code page // 4. For JIS X 0212, Remove the lead byte, & 0xFF7F, and use the CP20936 table to convert. // // Regarding Normalization: // Forms KC & KD are precluded because of things like halfwidth Katakana that has compatibility mappings // Form D is precluded because of 0x00a8, which changes to space + dierises. // // I think that IsAlwaysNormalized should probably return true for form C (but not certain) // // NOTE: We don't use JIS 0212 so we are basically a DBCS code page, we just have to modify // the 932 table we're basing this on. // using System; [Serializable()] internal class EUCJPEncoding : DBCSCodePageEncoding { // This pretends to be CP 932 as far as memory tables are concerned. public EUCJPEncoding() : base(51932, 932) { this.m_bUseMlangTypeForSerialization = true; } protected unsafe override String GetMemorySectionName() { int iUseCodePage = this.bFlagDataTable ? dataTableCodePage : CodePage; String strName = String.Format(CultureInfo.InvariantCulture, "CodePage_{0}_{1}_{2}_{3}_{4}_EUCJP", iUseCodePage, this.pCodePage->VersionMajor, this.pCodePage->VersionMinor, this.pCodePage->VersionRevision, this.pCodePage->VersionBuild); return strName; } // Clean up characters for EUC-JP code pages, etc. protected override bool CleanUpBytes(ref int bytes) { if (bytes >= 0x100) { // map extended char (0xfa40-0xfc4b) to a special range // (ported from mlang) if (bytes >= 0xfa40 && bytes <= 0xfc4b) { if ( bytes >= 0xfa40 && bytes <= 0xfa5b ) { if ( bytes <= 0xfa49 ) bytes = bytes - 0x0b51 ; else if ( bytes >= 0xfa4a && bytes <= 0xfa53 ) bytes = bytes - 0x072f6 ; else if ( bytes >= 0xfa54 && bytes <= 0xfa57 ) bytes = bytes - 0x0b5b ; else if ( bytes == 0xfa58 ) bytes = 0x878a ; else if ( bytes == 0xfa59 ) bytes = 0x8782 ; else if ( bytes == 0xfa5a ) bytes = 0x8784 ; else if ( bytes == 0xfa5b ) bytes = 0x879a ; } else if ( bytes >= 0xfa5c && bytes <= 0xfc4b ) { byte tc = unchecked((byte)bytes); if ( tc < 0x5c ) bytes = bytes - 0x0d5f; else if ( tc >= 0x80 && tc <= 0x9B ) bytes = bytes - 0x0d1d; else bytes = bytes - 0x0d1c; } } // Convert 932 code page to 20932 like code page range // (also ported from mlang) byte bLead = unchecked((byte)(bytes >> 8)); byte bTrail = unchecked((byte)bytes); bLead -= ((bLead > (byte)0x9f) ? (byte)0xb1 : (byte)0x71); bLead = (byte)((bLead << 1) + 1); if (bTrail > (byte)0x9e) { bTrail -= (byte)0x7e; bLead++; } else { if (bTrail > (byte)0x7e) bTrail--; bTrail -= (byte)0x1f; } bytes = ((int)bLead) << 8 | (int)bTrail | 0x8080; // // Don't step on our katakana special hack plane, if katakana space return false. // if (bytes >= 0x8E00 && bytes <= 0x8EFF) // return false; // Don't step out of our allocated lead byte area. // All DBCS lead and trail bytes should be >= 0xa1 and <= 0xfe if ((bytes & 0xFF00) < 0xa100 || (bytes & 0xFF00) > 0xfe00 || (bytes & 0xFF) < 0xa1 || (bytes & 0xFF) > 0xfe) return false; // WARNING: Our funky mapping allows illegal values, which we continue to use // so that we're compatible with Everett. } else { // For 51932 1/2 Katakana gets a 0x8E lead byte // Adjust 1/2 Katakana if (bytes >= 0xa1 && bytes <= 0xdf) { bytes |= 0x8E00; return true; } // 0x81-0x9f and 0xe0-0xfc CP 932 // 0x8e and 0xa1-0xfe CP 20932 (we don't use 8e though) // b0-df is 1/2 Katakana // So 81-9f & e0-fc are 932 lead bytes, a1-fe are our lead bytes // so ignore everything above 0x80 except 0xa0 and 0xff if (bytes >= 0x81 && bytes != 0xa0 && bytes != 0xff) { // We set diffent lead bytes later, so just return false return false; } } return true; } protected override unsafe void CleanUpEndBytes(char* chars) { // Need to special case CP 51932 // 0x81-0x9f and 0xe0-0xfc CP 932 // 0x8e and 0xa1-0xfe CP 20932 // 0x10 and 0x21-0x9? Us (remapping 932) // b0-df is 1/2 Katakana (trail byte) // A1-FE are DBCS code points for (int i = 0xA1; i <= 0xFE; i++) chars[i] = LEAD_BYTE_CHAR; // And 8E is lead byte for Katakana (already set) chars[0x8e] = LEAD_BYTE_CHAR; } } } // File provided for Reference Use Only by Microsoft Corporation (c) 2007.
Link Menu
This book is available now!
Buy at Amazon US or
Buy at Amazon UK
- SafeNativeMemoryHandle.cs
- MimeParameter.cs
- WSDualHttpSecurity.cs
- WindowsIPAddress.cs
- ImmutableObjectAttribute.cs
- ResolveMatchesCD1.cs
- ArglessEventHandlerProxy.cs
- Floater.cs
- GlobalItem.cs
- TreeViewCancelEvent.cs
- ColumnBinding.cs
- SynchronizedReadOnlyCollection.cs
- ConfigXmlAttribute.cs
- ThemeConfigurationDialog.cs
- RectangleConverter.cs
- DataRelationCollection.cs
- Point3DAnimation.cs
- SystemIcmpV6Statistics.cs
- RedistVersionInfo.cs
- ReadOnlyTernaryTree.cs
- LogicalMethodInfo.cs
- DoubleAnimation.cs
- PersonalizationProviderHelper.cs
- SessionChannels.cs
- CookielessData.cs
- RuntimeHelpers.cs
- SecurityManager.cs
- ToolStripGripRenderEventArgs.cs
- Vector3DKeyFrameCollection.cs
- ThemeConfigurationDialog.cs
- TextTreeNode.cs
- CannotUnloadAppDomainException.cs
- MemberInfoSerializationHolder.cs
- ChildrenQuery.cs
- AssociationType.cs
- CaseStatementProjectedSlot.cs
- GraphicsPath.cs
- ReversePositionQuery.cs
- ViewStateException.cs
- EntityFunctions.cs
- ScriptMethodAttribute.cs
- FigureParaClient.cs
- NotFiniteNumberException.cs
- EntityDataSourceStatementEditor.cs
- XmlQualifiedNameTest.cs
- Canvas.cs
- DrawingDrawingContext.cs
- CustomAttribute.cs
- BuildManager.cs
- QueueException.cs
- ShapeTypeface.cs
- EntityDataSourceContainerNameItem.cs
- BindingFormattingDialog.cs
- StringAnimationBase.cs
- RelationshipNavigation.cs
- GeometryValueSerializer.cs
- ClientWindowsAuthenticationMembershipProvider.cs
- Attributes.cs
- HyperLink.cs
- XpsS0ValidatingLoader.cs
- ObjectStateFormatter.cs
- XmlObjectSerializerWriteContextComplexJson.cs
- BamlLocalizableResource.cs
- TcpHostedTransportConfiguration.cs
- ListViewUpdateEventArgs.cs
- OleDbSchemaGuid.cs
- ExtendedPropertyCollection.cs
- MetabaseServerConfig.cs
- SubstitutionList.cs
- ImportCatalogPart.cs
- MembershipUser.cs
- NativeMethods.cs
- TrackingValidationObjectDictionary.cs
- ContainerUIElement3D.cs
- Point3DAnimationBase.cs
- Oid.cs
- HttpResponseInternalBase.cs
- WebPartTransformerCollection.cs
- CallSiteOps.cs
- ToolboxDataAttribute.cs
- TextElement.cs
- ColumnTypeConverter.cs
- ModuleConfigurationInfo.cs
- SerTrace.cs
- EncoderFallback.cs
- ServiceOperation.cs
- HandlerFactoryWrapper.cs
- Geometry.cs
- BlobPersonalizationState.cs
- TextRangeAdaptor.cs
- WriteFileContext.cs
- ToolboxItemFilterAttribute.cs
- CommandField.cs
- AttachedPropertyBrowsableForTypeAttribute.cs
- ObjectSet.cs
- JoinElimination.cs
- XmlSchemaSequence.cs
- RectConverter.cs
- OpenTypeLayoutCache.cs
- GeneralTransform3DGroup.cs