1.0 Overview
Fossil achieves efficient storage and low-bandwidth synchronization through the use of delta-compression. Instead of storing or transmitting the complete content of an artifact, Fossil stores or transmits only the changes relative to a related artifact.
This document describes the delta-encoding format used by Fossil. The intended audience is developers working on either Fossil itself, or on tools compatible with Fossil. Understanding of this document is not required for ordinary users of Fossil. This document is an implementation detail.
This document only describes the delta file format. A separate document describes one possible algorithm for generating deltas in this format.
1.1 Sample Software And Analysis Tools
The core routines used to generate and apply deltas in this format are contained in the delta.c source file. Interface logic, including "test-*" commands to directly manipulate deltas are contained in the deltacmd.c source file. SQL functions to create, apply, and analyze deltas are implemented by code in the deltafunc.c source file.
The following command-line tools are available to create and apply deltas and to test the delta logic:
- fossil test-delta → Run self-tests of the delta logic
- fossil test-delta-create X Y → compute a delta that converts file X into file Y. Output that delta.
- fossil test-delta-apply X D → apply delta D to input file X and output the result.
- fossil test-delta-analyze X Y → compute and delta that converts file X into file Y but instead of writing the delta to output, write performance information about the delta.
When running the fossil sql command to get an interactive SQL session connected to the repository, the following additional SQL functions are provided:
- delta_create(X,Y) → Compute a data that carries blob X into blob Y and return that delta as a blob.
- delta_apply(X,D) → Apply delta D to input blob X return a new blob which is the result.
- delta_output_size(D) → Return the size of the output that would result from applying delta D.
- delta_parse(D) → This is a table-valued function that returns one row for the header, for the trailer, and for each segment in delta D.
2.0 Structure
leftmargin = 0.1 box height 50% "Header" box same "Segments" box same "Trailer"
A delta consists of three parts, a "header", a "trailer", and a "segment-list" between them.
Both header and trailer provide information about the target helping the decoder, and the segment-list describes how the target can be constructed from the original.
2.1 Header
leftmargin = 0.1 box height 50% "Size" box same "\"\\n\""
The header consists of a single number followed by a newline character (ASCII 0x0a). The number is the length of the target in bytes.
This means that, given a delta, the decoder can compute the size of the target (and allocate any necessary memory based on that) by simply reading the first line of the delta and decoding the number found there. In other words, before it has to decode everything else.
2.2 Trailer
leftmargin = 0.1 box height 50% "Checksum" box same "\";\""
The trailer consists of a single number followed by a semicolon (ASCII 0x3b). This number is a checksum of the target and can be used by a decoder to verify that the delta applied correctly, reconstructing the target from the original.
The checksum is computed by treating the target as a series of 32-bit integer numbers (MSB first), and summing these up, modulo 2^32-1. A target whose length is not a multiple of 4 is padded with 0-bytes (ASCII 0x00) at the end.
By putting this information at the end of the delta a decoder has it available immediately after the target has been reconstructed fully.
2.3 Segment-List
leftmargin = 0.1 PART1: [ B1: box height 50% width 15% "" B2: box same "" B3: box same "" "***" box height 50% width 15% "" I1: line down 50% from B2 .s arrow right until even with B3.e box "Insert Literal" height 50% line down 75% from I1 .s arrow right until even with B3.e box "Copy Range" height 50% ] down PART2: [ "" box "Length" height 50% right box "\":\"" same box "Bytes" same ] with .nw at previous.sw
The segment-list of a delta describes how to create the target from the original by a combination of inserting literal byte-sequences and copying ranges of bytes from the original. This is where the compression takes place, by encoding the large common parts of original and target in small copy instructions.
The target is constructed from beginning to end, with the data generated by each instruction appended after the data of all previous instructions, with no gaps.
2.3.1 Insert Literal
A literal is specified by two elements, the size of the literal in bytes, and the bytes of the literal itself.
leftmargin = 0.1 box "Length" height 50% box "\":\"" same box "Bytes" same
The length is written first, followed by a colon character (ASCII 0x3a), followed by the bytes of the literal.
2.3.2 Copy Range
A range to copy is specified by two numbers, the offset of the first byte in the original to copy, and the size of the range, in bytes. The size zero is special, its usage indicates that the range extends to the end of the original.
leftmargin = 0.1 box "Length" height 50% box "\"@\"" same box "Offset" same box "\",\"" same
The length is written first, followed by an "at" character (ASCII 0x40), then the offset, followed by a comma (ASCII 0x2c).
3.0 Encoding of integers
The format currently handles only 32 bit integer numbers. They are written base-64 encoded, MSB first, and without leading "0"-characters, except if they are significant (i.e. 0 => "0").
The base-64 encoding uses one character for each 6 bits of the integer to be encoded. The encoding characters are:
0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ_abcdefghijklmnopqrstuvwxyz~
The least significant 6 bits of the integer are encoded by the first character, followed by the next 6 bits, and so on until all non-zero bits of the integer are encoded. The minimum number of encoding characters is used. Note that for integers less than 10, the base-64 coding is a ASCII decimal rendering of the number itself.
4.0 Examples
4.1 Integer encoding
Value | Encoding |
---|---|
0 | 0 |
6246 | 1Xb |
-1101438770 | 2zMM3E |
4.2 Delta encoding
An example of a delta using the specified encoding is:
1Xb 4E@0,2:thFN@4C,6:scenda1B@Jd,6:scenda5x@Kt,6:pieces79@Qt,F: Example: eskil~E@Y0,2zMM3E;
This can be taken apart into the following parts:
What | Encoding | Meaning | Details |
---|---|---|---|
Header | 1Xb | Size | 6246 |
S-List | 4E@0, | Copy | 270 @ 0 |
2:th | Literal | 2 'th' | |
FN@4C, | Copy | 983 @ 268 | |
6:scenda | Literal | 6 'scenda' | |
1B@Jd, | Copy | 75 @ 1256 | |
6:scenda | Literal | 6 'scenda' | |
5x@Kt, | Copy | 380 @ 1336 | |
6:pieces | Literal | 6 'pieces' | |
79@Qt, | Copy | 457 @ 1720 | |
F: Example: eskil | Literal | 15 ' Example: eskil' | |
~E@Y0, | Copy | 4046 @ 2176 | |
Trailer | 2zMM3E | Checksum | -1101438770 |
The unified diff behind the above delta is
bluepeak:(761) ~/Projects/Tcl/Fossil/Devel/devel > diff -u ../DELTA/old ../DELTA/new --- ../DELTA/old 2007-08-23 21:14:40.000000000 -0700 +++ ../DELTA/new 2007-08-23 21:14:33.000000000 -0700 @@ -5,7 +5,7 @@ * If the server does not have write permission on the database file, or on the directory containing the database file (and - it is thus unable to update database because it cannot create + it is thus unable to update the database because it cannot create a rollback journal) then it currently fails silently on a push. It needs to return a helpful error. @@ -27,8 +27,8 @@ * Additional information displayed for the "vinfo" page: + All leaves of this version that are not included in the - descendant list. With date, user, comment, and hyperlink. - Leaves in the descendant table should be marked as such. + descendant list. With date, user, comment, and hyperlink. + Leaves in the descendant table should be marked as such. See the compute_leaves() function to see how to find all leaves. + Add file diff links to the file change list. @@ -37,7 +37,7 @@ * The /xfer handler (for push, pull, and clone) does not do delta compression. This results in excess bandwidth usage. - There are some code in xfer.c that are sketches of ideas on + There are some pieces in xfer.c that are sketches of ideas on how to do delta compression, but nothing has been implemented. * Enhancements to the diff and tkdiff commands in the cli. @@ -45,7 +45,7 @@ single file. Allow diffs against any two arbitrary versions, not just diffs against the current check-out. Allow configuration options to replace tkdiff with some other - visual differ of the users choice. + visual differ of the users choice. Example: eskil. * Ticketing interface (expand this bullet)
Notes
- Pure text files generate a pure text delta.
- Binary files generate a delta that may contain some binary data.
- The delta encoding does not attempt to compress the content. It was considered to be much more sensible to do compression using a separate general-purpose compression library, like zlib.