zydronium/screwturn-4
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Page and message indexing is now more robust and pages that cannot be indexed are skipped.

Browse source
This commit is contained in:
Dario Solera 2010-05-30 06:29:13 +00:00
parent 4a91270ff1
commit 44311ba307
4 changed files with 79 additions and 55 deletions
Download patch file Download diff file Expand all files Collapse all files

4
AssemblyVersion.cs
View file

@ -16,5 +16,5 @@ using System.Reflection;
//
// You can specify all the values or you can default the Revision and Build Numbers
// by using the '*' as shown below:
[assembly: AssemblyVersion("3.0.2.540")]
[assembly: AssemblyFileVersion("3.0.2.540")]
[assembly: AssemblyVersion("3.0.2.541")]
[assembly: AssemblyFileVersion("3.0.2.541")]

64
Core/PagesStorageProvider.cs
View file

@ -1188,22 +1188,28 @@ namespace ScrewTurn.Wiki {
/// <returns>The number of indexed words, including duplicates.</returns>
private int IndexPage(PageContent content) {
lock(this) {
string documentName = PageDocument.GetDocumentName(content.PageInfo);
try {
string documentName = PageDocument.GetDocumentName(content.PageInfo);
DumpedDocument ddoc = new DumpedDocument(0, documentName, host.PrepareTitleForIndexing(content.PageInfo, content.Title),
PageDocument.StandardTypeTag, content.LastModified);
DumpedDocument ddoc = new DumpedDocument(0, documentName, host.PrepareTitleForIndexing(content.PageInfo, content.Title),
PageDocument.StandardTypeTag, content.LastModified);
// Store the document
// The content should always be prepared using IHost.PrepareForSearchEngineIndexing()
int count = index.StoreDocument(new PageDocument(content.PageInfo, ddoc, TokenizeContent),
content.Keywords, host.PrepareContentForIndexing(content.PageInfo, content.Content), null);
// Store the document
// The content should always be prepared using IHost.PrepareForSearchEngineIndexing()
int count = index.StoreDocument(new PageDocument(content.PageInfo, ddoc, TokenizeContent),
content.Keywords, host.PrepareContentForIndexing(content.PageInfo, content.Content), null);
if(count == 0 && content.Content.Length > 0) {
host.LogEntry("Indexed 0 words for page " + content.PageInfo.FullName + ": possible index corruption. Please report this error to the developers",
LogEntryType.Warning, null, this);
if(count == 0 && content.Content.Length > 0) {
host.LogEntry("Indexed 0 words for page " + content.PageInfo.FullName + ": possible index corruption. Please report this error to the developers",
LogEntryType.Warning, null, this);
}
return count;
}
catch(Exception ex) {
host.LogEntry("Page indexing error for " + content.PageInfo.FullName + " (skipping page): " + ex.ToString(), LogEntryType.Error, null, this);
return 0;
}
return count;
}
}
@ -2389,25 +2395,31 @@ namespace ScrewTurn.Wiki {
/// <returns>The number of indexed words, including duplicates.</returns>
private int IndexMessage(PageInfo page, int id, string subject, DateTime dateTime, string body) {
lock(this) {
// Trim "RE:" to avoid polluting the search engine index
if(subject.ToLowerInvariant().StartsWith("re:") && subject.Length > 3) subject = subject.Substring(3).Trim();
try {
// Trim "RE:" to avoid polluting the search engine index
if(subject.ToLowerInvariant().StartsWith("re:") && subject.Length > 3) subject = subject.Substring(3).Trim();
string documentName = MessageDocument.GetDocumentName(page, id);
string documentName = MessageDocument.GetDocumentName(page, id);
DumpedDocument ddoc = new DumpedDocument(0, documentName, host.PrepareTitleForIndexing(null, subject),
MessageDocument.StandardTypeTag, dateTime);
DumpedDocument ddoc = new DumpedDocument(0, documentName, host.PrepareTitleForIndexing(null, subject),
MessageDocument.StandardTypeTag, dateTime);
// Store the document
// The content should always be prepared using IHost.PrepareForSearchEngineIndexing()
int count = index.StoreDocument(new MessageDocument(page, id, ddoc, TokenizeContent), null,
host.PrepareContentForIndexing(null, body), null);
// Store the document
// The content should always be prepared using IHost.PrepareForSearchEngineIndexing()
int count = index.StoreDocument(new MessageDocument(page, id, ddoc, TokenizeContent), null,
host.PrepareContentForIndexing(null, body), null);
if(count == 0 && body.Length > 0) {
host.LogEntry("Indexed 0 words for message " + page.FullName + ":" + id.ToString() + ": possible index corruption. Please report this error to the developers",
LogEntryType.Warning, null, this);
if(count == 0 && body.Length > 0) {
host.LogEntry("Indexed 0 words for message " + page.FullName + ":" + id.ToString() + ": possible index corruption. Please report this error to the developers",
LogEntryType.Warning, null, this);
}
return count;
}
catch(Exception ex) {
host.LogEntry("Message indexing error for " + page.FullName + ":" + id.ToString() + " (skipping message): " + ex.ToString(), LogEntryType.Error, null, this);
return 0;
}
return count;
}
}

64
SqlProvidersCommon/SqlPagesStorageProviderBase.cs
View file

@ -592,22 +592,28 @@ namespace ScrewTurn.Wiki.Plugins.SqlCommon {
/// <param name="transaction">The current transaction.</param>
/// <returns>The number of indexed words, including duplicates.</returns>
private int IndexPage(PageContent content, DbTransaction transaction) {
string documentName = PageDocument.GetDocumentName(content.PageInfo);
try {
string documentName = PageDocument.GetDocumentName(content.PageInfo);
DumpedDocument ddoc = new DumpedDocument(0, documentName, host.PrepareTitleForIndexing(content.PageInfo, content.Title),
PageDocument.StandardTypeTag, content.LastModified);
DumpedDocument ddoc = new DumpedDocument(0, documentName, host.PrepareTitleForIndexing(content.PageInfo, content.Title),
PageDocument.StandardTypeTag, content.LastModified);
// Store the document
// The content should always be prepared using IHost.PrepareForSearchEngineIndexing()
int count = index.StoreDocument(new PageDocument(content.PageInfo, ddoc, TokenizeContent),
content.Keywords, host.PrepareContentForIndexing(content.PageInfo, content.Content), transaction);
// Store the document
// The content should always be prepared using IHost.PrepareForSearchEngineIndexing()
int count = index.StoreDocument(new PageDocument(content.PageInfo, ddoc, TokenizeContent),
content.Keywords, host.PrepareContentForIndexing(content.PageInfo, content.Content), transaction);
if(count == 0 && content.Content.Length > 0) {
host.LogEntry("Indexed 0 words for page " + content.PageInfo.FullName + ": possible index corruption. Please report this error to the developers",
LogEntryType.Warning, null, this);
if(count == 0 && content.Content.Length > 0) {
host.LogEntry("Indexed 0 words for page " + content.PageInfo.FullName + ": possible index corruption. Please report this error to the developers",
LogEntryType.Warning, null, this);
}
return count;
}
catch(Exception ex) {
host.LogEntry("Page indexing error for " + content.PageInfo.FullName + " (skipping page): " + ex.ToString(), LogEntryType.Error, null, this);
return 0;
}
return count;
}
/// <summary>
@ -647,25 +653,31 @@ namespace ScrewTurn.Wiki.Plugins.SqlCommon {
/// <param name="transaction">The current transaction.</param>
/// <returns>The number of indexed words, including duplicates.</returns>
private int IndexMessage(PageInfo page, int id, string subject, DateTime dateTime, string body, DbTransaction transaction) {
// Trim "RE:" to avoid polluting the search engine index
if(subject.ToLowerInvariant().StartsWith("re:") && subject.Length > 3) subject = subject.Substring(3).Trim();
try {
// Trim "RE:" to avoid polluting the search engine index
if(subject.ToLowerInvariant().StartsWith("re:") && subject.Length > 3) subject = subject.Substring(3).Trim();
string documentName = MessageDocument.GetDocumentName(page, id);
string documentName = MessageDocument.GetDocumentName(page, id);
DumpedDocument ddoc = new DumpedDocument(0, documentName, host.PrepareTitleForIndexing(null, subject),
MessageDocument.StandardTypeTag, dateTime);
DumpedDocument ddoc = new DumpedDocument(0, documentName, host.PrepareTitleForIndexing(null, subject),
MessageDocument.StandardTypeTag, dateTime);
// Store the document
// The content should always be prepared using IHost.PrepareForSearchEngineIndexing()
int count = index.StoreDocument(new MessageDocument(page, id, ddoc, TokenizeContent), null,
host.PrepareContentForIndexing(null, body), transaction);
// Store the document
// The content should always be prepared using IHost.PrepareForSearchEngineIndexing()
int count = index.StoreDocument(new MessageDocument(page, id, ddoc, TokenizeContent), null,
host.PrepareContentForIndexing(null, body), transaction);
if(count == 0 && body.Length > 0) {
host.LogEntry("Indexed 0 words for message " + page.FullName + ":" + id.ToString() + ": possible index corruption. Please report this error to the developers",
LogEntryType.Warning, null, this);
if(count == 0 && body.Length > 0) {
host.LogEntry("Indexed 0 words for message " + page.FullName + ":" + id.ToString() + ": possible index corruption. Please report this error to the developers",
LogEntryType.Warning, null, this);
}
return count;
}
catch(Exception ex) {
host.LogEntry("Message indexing error for " + page.FullName + ":" + id.ToString() + " (skipping message): " + ex.ToString(), LogEntryType.Error, null, this);
return 0;
}
return count;
}
/// <summary>

2
SqlServerProviders/SqlServerPagesStorageProvider.cs
View file

@ -13,7 +13,7 @@ namespace ScrewTurn.Wiki.Plugins.SqlServer {
/// </summary>
public class SqlServerPagesStorageProvider : SqlPagesStorageProviderBase, IPagesStorageProviderV30 {
private readonly ComponentInformation info = new ComponentInformation("SQL Server Pages Storage Provider", "Threeplicate Srl", "3.0.1.471", "http://www.screwturn.eu", "http://www.screwturn.eu/Version/SQLServerProv/Pages.txt");
private readonly ComponentInformation info = new ComponentInformation("SQL Server Pages Storage Provider", "Threeplicate Srl", "3.0.1.541", "http://www.screwturn.eu", "http://www.screwturn.eu/Version/SQLServerProv/Pages.txt");
private readonly SqlServerCommandBuilder commandBuilder = new SqlServerCommandBuilder();