Skip to content

v0.1.0

Latest

Choose a tag to compare

@github-actions github-actions released this 14 Nov 01:28
· 51 commits to main since this release

First release of TdbIO.

Added

  • TdbFile.DetectKind and TdbFile.Open for paths and streams: the kind of a file is found from its content (reports by
    their .xsr name), and the file is read into a typed, read-only document.
  • Gzip unpacking with checks of the header, the size and the CRC-32, capped by TdbReadOptions.MaxUnpackedLength
    (1 GiB by default).
  • Model databases (.db1): header, id map, text sections and chains, save stamp dates (TextChain.SavedAt), and the
    table sections with a row view that stores no row objects (Db1Document.ReadTables).
  • Model records files (.db2), DBV@ files, id range files (.idrm), guid.mapper and the header of history.db.
  • DSTV NC files (.nc1): the ST header fields and the blocks, with AK and BO lines parsed.
  • Templates (.tpl, .rpt) as a tree of blocks and properties, and reports (.xsr) as lines and pages, with encoding
    detection (ASCII, UTF-8, code pages 1251 and 1252) or a forced encoding.
  • InvalidDataException with the offset of the problem in the message and in Data["Offset"], and
    TdbDocument.Warnings for values that are kept although they differ from what is expected.
  • The tdbio command line tool (package TdbIO.Cli) with the commands info, tables and text, text or JSON
    output, and exit codes 0, 1 and 2.
  • Targets .NET 8 and .NET 10, with no package dependencies. The library never writes to the console.

Not yet supported

  • The content of table rows and .db2 rows is given as raw bytes; it is not decoded into model objects.
  • Drawing files are recognised but not read. Of history.db only the header is read.
  • Gzip data with more than one member, with bytes after the compressed data or with header flags, and UTF-16 templates
    and reports, are rejected.