use 'best' hash for source authentication

Collect all hashes we can get from the source record and put them into a
HashStringList so that 'apt-get source' can use it instead of using
always the MD5sum.

We therefore also deprecate the MD5 struct member in favor of the list.

While at it, the parsing of the Files is enhanced so that records which
miss "Files" (aka MD5 checksums) are still searched for other checksums
as they include just as much data, just not with a nice and catchy name.

This is a cherry-pick of 1262d35 with some dirty tricks to preserve ABI.

LP: 1098738

4 files changed, 167 insertions, 48 deletions

diff --git a/apt-pkg/deb/debsrcrecords.cc b/apt-pkg/deb/debsrcrecords.cc
index a444cbe4d..49a348dd4 100644
--- a/apt-pkg/deb/debsrcrecords.cc
+++ b/apt-pkg/deb/debsrcrecords.cc
@@ -118,13 +118,32 @@ bool debSrcRecordParser::BuildDepends(std::vector<pkgSrcRecords::Parser::BuildDe
 // ---------------------------------------------------------------------
 /* This parses the list of files and returns it, each file is required to have
    a complete source package */
-bool debSrcRecordParser::Files(std::vector<pkgSrcRecords::File> &List)
+bool debSrcRecordParser::Files(std::vector<pkgSrcRecords::File> &F)
 {
-   List.erase(List.begin(),List.end());
-   
-   string Files = Sect.FindS("Files");
-   if (Files.empty() == true)
+   std::vector<pkgSrcRecords::File2> F2;
+   if (Files2(F2) == false)
       return false;
+   for (std::vector<pkgSrcRecords::File2>::const_iterator f2 = F2.begin(); f2 != F2.end(); ++f2)
+   {
+      pkgSrcRecords::File2 f;
+#if __GNUC__ >= 4
+	#pragma GCC diagnostic push
+	#pragma GCC diagnostic ignored "-Wdeprecated-declarations"
+#endif
+      f.MD5Hash = f2->MD5Hash;
+      f.Size = f2->Size;
+#if __GNUC__ >= 4
+	#pragma GCC diagnostic pop
+#endif
+      f.Path = f2->Path;
+      f.Type = f2->Type;
+      F.push_back(f);
+   }
+   return true;
+}
+bool debSrcRecordParser::Files2(std::vector<pkgSrcRecords::File2> &List)
+{
+   List.clear();
 
    // Stash the / terminated directory prefix
    string Base = Sect.FindS("Directory");
@@ -133,51 +152,106 @@ bool debSrcRecordParser::Files(std::vector<pkgSrcRecords::File> &List)
 
    std::vector<std::string> const compExts = APT::Configuration::getCompressorExtensions();
 
-   // Iterate over the entire list grabbing each triplet
-   const char *C = Files.c_str();
-   while (*C != 0)
-   {   
-      pkgSrcRecords::File F;
-      string Size;
-      
-      // Parse each of the elements
-      if (ParseQuoteWord(C,F.MD5Hash) == false ||
-	  ParseQuoteWord(C,Size) == false ||
-	  ParseQuoteWord(C,F.Path) == false)
-	 return _error->Error("Error parsing file record");
-      
-      // Parse the size and append the directory
-      F.Size = atoi(Size.c_str());
-      F.Path = Base + F.Path;
-      
-      // Try to guess what sort of file it is we are getting.
-      string::size_type Pos = F.Path.length()-1;
-      while (1)
+   for (char const * const * type = HashString::SupportedHashes(); *type != NULL; ++type)
+   {
+      // derive field from checksum type
+      std::string checksumField("Checksums-");
+      if (strcmp(*type, "MD5Sum") == 0)
+	 checksumField = "Files"; // historic name for MD5 checksums
+      else
+	 checksumField.append(*type);
+
+      string const Files = Sect.FindS(checksumField.c_str());
+      if (Files.empty() == true)
+	 continue;
+
+      // Iterate over the entire list grabbing each triplet
+      const char *C = Files.c_str();
+      while (*C != 0)
       {
-	 string::size_type Tmp = F.Path.rfind('.',Pos);
-	 if (Tmp == string::npos)
-	    break;
-	 if (F.Type == "tar") {
-	    // source v3 has extension 'debian.tar.*' instead of 'diff.*'
-	    if (string(F.Path, Tmp+1, Pos-Tmp) == "debian")
-	       F.Type = "diff";
-	    break;
-	 }
-	 F.Type = string(F.Path,Tmp+1,Pos-Tmp);
-	 
-	 if (std::find(compExts.begin(), compExts.end(), std::string(".").append(F.Type)) != compExts.end() ||
-	     F.Type == "tar")
+	 string hash, size, path;
+
+	 // Parse each of the elements
+	 if (ParseQuoteWord(C, hash) == false ||
+	       ParseQuoteWord(C, size) == false ||
+	       ParseQuoteWord(C, path) == false)
+	    return _error->Error("Error parsing file record in %s of source package %s", checksumField.c_str(), Package().c_str());
+
+	 HashString const hashString(*type, hash);
+	 if (Base.empty() == false)
+	    path = Base + path;
+
+	 // look if we have a record for this file already
+	 std::vector<pkgSrcRecords::File2>::iterator file = List.begin();
+	 for (; file != List.end(); ++file)
+	    if (file->Path == path)
+	       break;
+
+	 // we have it already, store the new hash and be done
+	 if (file != List.end())
 	 {
-	    Pos = Tmp-1;
+#if __GNUC__ >= 4
+	// set for compatibility only, so warn users not us
+	#pragma GCC diagnostic push
+	#pragma GCC diagnostic ignored "-Wdeprecated-declarations"
+#endif
+	    if (checksumField == "Files")
+	       file->MD5Hash = hash;
+#if __GNUC__ >= 4
+	#pragma GCC diagnostic pop
+#endif
+	    // an error here indicates that we have two different hashes for the same file
+	    if (file->Hashes.push_back(hashString) == false)
+	       return _error->Error("Error parsing checksum in %s of source package %s", checksumField.c_str(), Package().c_str());
 	    continue;
 	 }
-	 
-	 break;
+
+	 // we haven't seen this file yet
+	 pkgSrcRecords::File2 F;
+	 F.Path = path;
+	 F.FileSize = strtoull(size.c_str(), NULL, 10);
+	 F.Hashes.push_back(hashString);
+
+#if __GNUC__ >= 4
+	// set for compatibility only, so warn users not us
+	#pragma GCC diagnostic push
+	#pragma GCC diagnostic ignored "-Wdeprecated-declarations"
+#endif
+	 F.Size = F.FileSize;
+	 if (checksumField == "Files")
+	    F.MD5Hash = hash;
+#if __GNUC__ >= 4
+	#pragma GCC diagnostic pop
+#endif
+
+	 // Try to guess what sort of file it is we are getting.
+	 string::size_type Pos = F.Path.length()-1;
+	 while (1)
+	 {
+	    string::size_type Tmp = F.Path.rfind('.',Pos);
+	    if (Tmp == string::npos)
+	       break;
+	    if (F.Type == "tar") {
+	       // source v3 has extension 'debian.tar.*' instead of 'diff.*'
+	       if (string(F.Path, Tmp+1, Pos-Tmp) == "debian")
+		  F.Type = "diff";
+	       break;
+	    }
+	    F.Type = string(F.Path,Tmp+1,Pos-Tmp);
+
+	    if (std::find(compExts.begin(), compExts.end(), std::string(".").append(F.Type)) != compExts.end() ||
+		  F.Type == "tar")
+	    {
+	       Pos = Tmp-1;
+	       continue;
+	    }
+
+	    break;
+	 }
+	 List.push_back(F);
       }
-      
-      List.push_back(F);
    }
-   
+
    return true;
 }
 									/*}}}*/
diff --git a/apt-pkg/deb/debsrcrecords.h b/apt-pkg/deb/debsrcrecords.h
index b65d1480b..2a3fc86c9 100644
--- a/apt-pkg/deb/debsrcrecords.h
+++ b/apt-pkg/deb/debsrcrecords.h
@@ -53,6 +53,7 @@ class debSrcRecordParser : public pkgSrcRecords::Parser
       return std::string(Start,Stop);
    };
    virtual bool Files(std::vector<pkgSrcRecords::File> &F);
+   bool Files2(std::vector<pkgSrcRecords::File2> &F);
 
    debSrcRecordParser(std::string const &File,pkgIndexFile const *Index) 
       : Parser(Index), Fd(File,FileFd::ReadOnly, FileFd::Extension), Tags(&Fd,102400),
diff --git a/apt-pkg/srcrecords.cc b/apt-pkg/srcrecords.cc
index 81b1c545d..3175ee75f 100644
--- a/apt-pkg/srcrecords.cc
+++ b/apt-pkg/srcrecords.cc
@@ -14,6 +14,7 @@
 #include<config.h>
 
 #include <apt-pkg/srcrecords.h>
+#include <apt-pkg/debsrcrecords.h>
 #include <apt-pkg/error.h>
 #include <apt-pkg/sourcelist.h>
 #include <apt-pkg/metaindex.h>
@@ -147,5 +148,33 @@ const char *pkgSrcRecords::Parser::BuildDepType(unsigned char const &Type)
    return fields[Type];
 }
 									/*}}}*/
+bool pkgSrcRecords::Parser::Files2(std::vector<pkgSrcRecords::File2> &F2)/*{{{*/
+{
+   debSrcRecordParser * const deb = dynamic_cast<debSrcRecordParser*>(this);
+   if (deb != NULL)
+      return deb->Files2(F2);
 
-
+   std::vector<pkgSrcRecords::File> F;
+   if (Files(F) == false)
+      return false;
+   for (std::vector<pkgSrcRecords::File>::const_iterator f = F.begin(); f != F.end(); ++f)
+   {
+      pkgSrcRecords::File2 f2;
+#if __GNUC__ >= 4
+	#pragma GCC diagnostic push
+	#pragma GCC diagnostic ignored "-Wdeprecated-declarations"
+#endif
+      f2.MD5Hash = f->MD5Hash;
+      f2.Size = f->Size;
+      f2.Hashes.push_back(HashString("MD5Sum", f->MD5Hash));
+      f2.FileSize = f->Size;
+#if __GNUC__ >= 4
+	#pragma GCC diagnostic pop
+#endif
+      f2.Path = f->Path;
+      f2.Type = f->Type;
+      F2.push_back(f2);
+   }
+   return true;
+}
+									/*}}}*/
diff --git a/apt-pkg/srcrecords.h b/apt-pkg/srcrecords.h
index e000e176a..dde22bd65 100644
--- a/apt-pkg/srcrecords.h
+++ b/apt-pkg/srcrecords.h
@@ -14,6 +14,7 @@
 #define PKGLIB_SRCRECORDS_H
 
 #include <apt-pkg/macros.h>
+#include <apt-pkg/hashes.h>
 
 #include <string>
 #include <vector>
@@ -29,15 +30,28 @@ class pkgSrcRecords
 {
    public:
 
+#if __GNUC__ >= 4
+	// ensure that con- & de-structor don't trigger this warning
+	#pragma GCC diagnostic push
+	#pragma GCC diagnostic ignored "-Wdeprecated-declarations"
+#endif
    // Describes a single file
    struct File
    {
-      std::string MD5Hash;
-      unsigned long Size;
+      APT_DEPRECATED std::string MD5Hash;
+      APT_DEPRECATED unsigned long Size;
       std::string Path;
       std::string Type;
    };
-   
+   struct File2 : public File
+   {
+      unsigned long long FileSize;
+      HashStringList Hashes;
+   };
+#if __GNUC__ >= 4
+	#pragma GCC diagnostic pop
+#endif
+
    // Abstract parser for each source record
    class Parser
    {
@@ -77,6 +91,7 @@ class pkgSrcRecords
       static const char *BuildDepType(unsigned char const &Type) APT_PURE;
 
       virtual bool Files(std::vector<pkgSrcRecords::File> &F) = 0;
+      bool Files2(std::vector<pkgSrcRecords::File2> &F);
       
       Parser(const pkgIndexFile *Index) : iIndex(Index) {};
       virtual ~Parser() {};