Add DBF ingestion and typed CSV/XLSX export

.github/workflows/dotnet-tests.yml

@@ -35,6 +35,40 @@ env:
   SUMMARIZE_FAILURES: true # set to 'false' to disable summarizing failing tests
 
 jobs:
+  dbf-contracts:
+    name: 'DBF contracts / ${{ matrix.os }} / ${{ matrix.framework }}'
+    runs-on: ${{ matrix.os }}
+    timeout-minutes: 15
+    strategy:
+      fail-fast: false
+      matrix:
+        os: [ubuntu-latest, windows-latest]
+        framework: [net8.0, net10.0]
+        include:
+          - os: windows-latest
+            framework: net472
+    steps:
+      - uses: actions/checkout@v7
+        with:
+          persist-credentials: false
+      - uses: actions/setup-dotnet@v6
+        with:
+          dotnet-version: ${{ env.DOTNET_VERSION }}
+      - name: Read independent DBF fixtures and reopen CSV/XLSX output
+        shell: pwsh
+        run: |
+          $ErrorActionPreference = 'Stop'
+          $PSNativeCommandUseErrorActionPreference = $true
+          dotnet test OfficeIMO.Reader.Dbf.Tests/OfficeIMO.Reader.Dbf.Tests.csproj --configuration ${{ env.BUILD_CONFIGURATION }} --framework ${{ matrix.framework }} --filter 'Category!=Performance' --logger 'trx;LogFileName=dbf.trx' --results-directory TestResults/Dbf
+      - name: Upload test results
+        if: always()
+        uses: actions/upload-artifact@v7
+        with:
+          name: dbf-tests-${{ matrix.os }}-${{ matrix.framework }}
+          path: TestResults/Dbf
+          if-no-files-found: ignore
+          retention-days: 7
+
   browser-exports:
     name: 'Browser exports / Chromium, Firefox and WebKit'
     runs-on: windows-latest

Build/CompatibilityCatalog/ConversionApiCompileContract.cs

@@ -1,4 +1,5 @@
 using OfficeIMO.Adf;
+using DBAClientX.Dbf;
 using OfficeIMO.AsciiDoc;
 using OfficeIMO.Confluence;
 using OfficeIMO.CSV;
@@ -165,6 +166,13 @@ internal static void Verify(
         _ = ConfluenceContentConverter.ToHtml(confluencePage);
         _ = ExcelDocumentCsvExtensions.ToExcelDocument(csv);
         _ = ExcelSheetCsvExtensions.ToCsv(excel.Sheets[0]);
+        using (DbfDataReader dbf = DbfDataReader.Open(source))
+        using (StringWriter csvOutput = new StringWriter()) {
+            CsvDocument.WriteDataReader(csvOutput, dbf);
+        }
+        using (DbfDataReader dbf = DbfDataReader.Open(source)) {
+            _ = ExcelDocument.WriteDataReader(stream, dbf);
+        }
         _ = OfficeIMO.Markup.Word.OfficeMarkupWordConverterExtensions.ToWordDocumentResult(officeMarkup);
         _ = OfficeIMO.Markup.Excel.OfficeMarkupExcelConverterExtensions.ToExcelDocumentResult(officeMarkup);
         _ = OfficeIMO.Markup.PowerPoint.OfficeMarkupPowerPointConverterExtensions.ToPowerPointPresentationResult(officeMarkup);

Build/CompatibilityCatalog/OfficeIMO.CompatibilityCatalog.Tool.csproj

@@ -7,6 +7,7 @@
     <IsPackable>false</IsPackable>
   </PropertyGroup>
   <ItemGroup>
+    <PackageReference Include="DBAClientX.Dbf" Version="1.0.0" />
     <ProjectReference Include="../../OfficeIMO.OpenDocument.Odg.Pdf/OfficeIMO.OpenDocument.Odg.Pdf.csproj" />
     <ProjectReference Include="../../OfficeIMO.Adf/OfficeIMO.Adf.csproj" />
     <ProjectReference Include="../../OfficeIMO.Confluence/OfficeIMO.Confluence.csproj" />

Build/Test-AotProjectCoverage.ps1

@@ -48,6 +48,11 @@ $nativeTools = @(
     }
 )
 $nonNativeValidated = @(
+    [ordered]@{
+        name = 'OfficeIMO.Reader.Dbf'
+        classification = 'managed-cross-platform'
+        evidence = 'DBF ingestion is qualified through managed .NET 8 and 10 extraction and CSV/XLSX conversion tests; the DBAClientX.Dbf owner and Reader adapter are not rooted in a NativeAOT qualification host.'
+    }
     [ordered]@{
         name = 'OfficeIMO.OpenDocument.Odg.Pdf'
         classification = 'managed-cross-platform'

Build/project.build.json

@@ -91,6 +91,7 @@
     "OfficeIMO.Reader.All": "3.4.X",
     "OfficeIMO.Reader.AsciiDoc": "3.4.X",
     "OfficeIMO.Reader.Csv": "3.4.X",
+    "OfficeIMO.Reader.Dbf": "3.4.X",
     "OfficeIMO.Reader.DocBook": "3.4.X",
     "OfficeIMO.Reader.Email": "3.4.X",
     "OfficeIMO.Reader.Excel": "3.4.X",

Docs/Compatibility/generated/README.md

@@ -23,7 +23,7 @@ dotnet run --framework net8.0 --project Build/CompatibilityCatalog/OfficeIMO.Com
 | OfficeIMO.Excel.LegacyXls | 1 | 28 | [JSON](excel-legacy-xls.json) | [Markdown](excel-legacy-xls.md) |
 | OfficeIMO.Excel.Xlsb | 1 | 20 | [JSON](excel-xlsb.json) | [Markdown](excel-xlsb.md) |
 | OfficeIMO.PowerPoint.LegacyPpt | 1 | 56 | [JSON](powerpoint-legacy-ppt.json) | [Markdown](powerpoint-legacy-ppt.md) |
-| OfficeIMO.Operations | 1 | 1385 | [JSON](package-operations.json) | [Markdown](package-operations.md) |
+| OfficeIMO.Operations | 1 | 1387 | [JSON](package-operations.json) | [Markdown](package-operations.md) |
 | OfficeIMO.Provenance | 1 | 11 | [JSON](provenance.json) | [Markdown](provenance.md) |
 | OfficeIMO.ProtectedContent | 2 | 18 | [JSON](protected-content.json) | [Markdown](protected-content.md) |
 

Docs/Compatibility/generated/conversion-routes.json

@@ -1481,6 +1481,46 @@
       "browserAvailable":false,
       "agentDiscoverable":true
     },
+    {
+      "id":"dbf-csv",
+      "source":"DBF/xBase",
+      "target":"CSV",
+      "inputKind":"File",
+      "sourceExtensions":[ ".dbf" ],
+      "targetExtension":".csv",
+      "packageId":"OfficeIMO.CSV",
+      "api":"DbfDataReader.Open(path), then CsvDocument.WriteDataReader(writer, reader, options)",
+      "description":"Export typed DBF rows and supported memos through DBAClientX.Dbf; binary values become Base64.",
+      "fidelity":"Semantic",
+      "textFormatting":"DataOnly",
+      "textFormattingContract":"DBF carries typed table values, not document typography or layout; binary values become Base64, and DBF indexes, deletion flags and native schema metadata are not exported.",
+      "supportLevel":"Targeted",
+      "supportEvidence":"Independent Python-produced dBASE III, FoxPro 2 and Visual FoxPro fixtures exercise typed rows, deleted-row filtering, DBT/FPT memos, binary Base64, CSV reopen, and streaming/buffered XLSX reopen and Open XML validation.",
+      "knownLimitations":"Requires DBAClientX.Dbf's bounded field profiles. Native application acceptance and Visual FoxPro 31/32 generations remain unqualified. Output omits DBF indexes, deletion flags and native schema metadata; CSV nulls need a selected marker, and XLSX follows Excel numeric precision and cell text limits.",
+      "resultContract":"void",
+      "browserAvailable":false,
+      "agentDiscoverable":true
+    },
+    {
+      "id":"dbf-xlsx",
+      "source":"DBF/xBase",
+      "target":"XLSX",
+      "inputKind":"File",
+      "sourceExtensions":[ ".dbf" ],
+      "targetExtension":".xlsx",
+      "packageId":"OfficeIMO.Excel",
+      "api":"DbfDataReader.Open(path), then ExcelDocument.WriteDataReader(stream, reader, options)",
+      "description":"Create a workbook from typed DBF rows and supported memos through DBAClientX.Dbf.",
+      "fidelity":"Editable",
+      "textFormatting":"DataOnly",
+      "textFormattingContract":"DBF carries typed table values, not document typography or layout; binary values become Base64, and DBF indexes, deletion flags and native schema metadata are not exported.",
+      "supportLevel":"Targeted",
+      "supportEvidence":"Independent Python-produced dBASE III, FoxPro 2 and Visual FoxPro fixtures exercise typed rows, deleted-row filtering, DBT/FPT memos, binary Base64, CSV reopen, and streaming/buffered XLSX reopen and Open XML validation.",
+      "knownLimitations":"Requires DBAClientX.Dbf's bounded field profiles. Native application acceptance and Visual FoxPro 31/32 generations remain unqualified. Output omits DBF indexes, deletion flags and native schema metadata; CSV nulls need a selected marker, and XLSX follows Excel numeric precision and cell text limits.",
+      "resultContract":"ExcelDataSetImportResult",
+      "browserAvailable":false,
+      "agentDiscoverable":true
+    },
     {
       "id":"officemarkup-docx",
       "source":"OfficeIMO Markup",

Docs/Compatibility/generated/conversion-routes.md

@@ -80,6 +80,8 @@ Schema version: 9
 | confluence-html | Confluence | HTML | OfficeIMO.Confluence | Semantic | SyntaxSubset | Preserves the text styling represented by the source and destination profiles; unsupported native decoration variants, arbitrary CSS presentation, and format-specific layout are reported or simplified. | Targeted | Materialized Confluence page and body fixtures exercise ADF and storage representations through the shared ADF, Markdown, and HTML converters with combined diagnostics. | This is content-body conversion, not live page synchronization; Confluence-only macros, layouts, extensions, and presentation outside the supported ADF/storage subset may be simplified. | No | `ConfluenceContentConverter.ToHtml(page)` | ConfluenceContentConversionResult<string> | Project a materialized Confluence page body to HTML with fidelity diagnostics. |
 | csv-xlsx | CSV | XLSX | OfficeIMO.Excel.Csv | Editable | DataOnly | CSV and TSV carry values, delimiters, and records only; font family, size, color, emphasis, decoration, scripts, casing metadata, and layout are intentionally not representable. | Established | Delimited-data fixtures cover streams, files, delimiter detection, typed cell import, worksheet-range export, reopen behavior, and cancellation. | CSV and TSV have no font, rich-text, formula, multi-sheet, drawing, or layout model; those workbook semantics are intentionally absent from CSV output. | No | `CsvDocument.Load(stream).ToExcelDocument(options)` | ExcelDocument | Import delimited values into an editable workbook. |
 | xlsx-csv | XLSX | CSV | OfficeIMO.Excel.Csv | Semantic | DataOnly | CSV and TSV carry values, delimiters, and records only; font family, size, color, emphasis, decoration, scripts, casing metadata, and layout are intentionally not representable. | Established | Delimited-data fixtures cover streams, files, delimiter detection, typed cell import, worksheet-range export, reopen behavior, and cancellation. | CSV and TSV have no font, rich-text, formula, multi-sheet, drawing, or layout model; those workbook semantics are intentionally absent from CSV output. | No | `ExcelDocument.Load(stream).Sheets[0].ToCsv(options)` | string | Export a worksheet used range as delimited values. |
+| dbf-csv | DBF/xBase | CSV | OfficeIMO.CSV | Semantic | DataOnly | DBF carries typed table values, not document typography or layout; binary values become Base64, and DBF indexes, deletion flags and native schema metadata are not exported. | Targeted | Independent Python-produced dBASE III, FoxPro 2 and Visual FoxPro fixtures exercise typed rows, deleted-row filtering, DBT/FPT memos, binary Base64, CSV reopen, and streaming/buffered XLSX reopen and Open XML validation. | Requires DBAClientX.Dbf's bounded field profiles. Native application acceptance and Visual FoxPro 31/32 generations remain unqualified. Output omits DBF indexes, deletion flags and native schema metadata; CSV nulls need a selected marker, and XLSX follows Excel numeric precision and cell text limits. | No | `DbfDataReader.Open(path), then CsvDocument.WriteDataReader(writer, reader, options)` | void | Export typed DBF rows and supported memos through DBAClientX.Dbf; binary values become Base64. |
+| dbf-xlsx | DBF/xBase | XLSX | OfficeIMO.Excel | Editable | DataOnly | DBF carries typed table values, not document typography or layout; binary values become Base64, and DBF indexes, deletion flags and native schema metadata are not exported. | Targeted | Independent Python-produced dBASE III, FoxPro 2 and Visual FoxPro fixtures exercise typed rows, deleted-row filtering, DBT/FPT memos, binary Base64, CSV reopen, and streaming/buffered XLSX reopen and Open XML validation. | Requires DBAClientX.Dbf's bounded field profiles. Native application acceptance and Visual FoxPro 31/32 generations remain unqualified. Output omits DBF indexes, deletion flags and native schema metadata; CSV nulls need a selected marker, and XLSX follows Excel numeric precision and cell text limits. | No | `DbfDataReader.Open(path), then ExcelDocument.WriteDataReader(stream, reader, options)` | ExcelDataSetImportResult | Create a workbook from typed DBF rows and supported memos through DBAClientX.Dbf. |
 | officemarkup-docx | OfficeIMO Markup | DOCX | OfficeIMO.Markup.Word | Editable | EditableEquivalent | Authors editable native typography, including family, size, color, emphasis, decoration, script, and casing, in the generated Office document; diagnostics identify profile-specific approximations and omissions. | Established | Typed profile fixtures verify parser diagnostics, editable destination artifacts, font family/size/color, bold, italic, underline variants, strike, scripts, case transforms, small caps where native, and target-specific layout. | OfficeIMO Markup is a directed authoring format rather than a lossless Office round trip; unsupported blocks and destination-only effects are diagnosed, simplified, or omitted according to exporter options. | No | `OfficeMarkupParser.Parse(markup, options).Document.ToWordDocumentResult(exportOptions)` | OfficeMarkupConversionResult<WordDocument> | Render document-profile OfficeIMO Markup into an editable Word document. |
 | officemarkup-xlsx | OfficeIMO Markup | XLSX | OfficeIMO.Markup.Excel | Editable | EditableEquivalent | Authors editable native typography, including family, size, color, emphasis, decoration, script, and casing, in the generated Office document; diagnostics identify profile-specific approximations and omissions. | Established | Typed profile fixtures verify parser diagnostics, editable destination artifacts, font family/size/color, bold, italic, underline variants, strike, scripts, case transforms, small caps where native, and target-specific layout. | OfficeIMO Markup is a directed authoring format rather than a lossless Office round trip; unsupported blocks and destination-only effects are diagnosed, simplified, or omitted according to exporter options. | No | `OfficeMarkupParser.Parse(markup, options).Document.ToExcelDocumentResult(exportOptions)` | OfficeMarkupConversionResult<ExcelDocument> | Render workbook-profile OfficeIMO Markup into an editable Excel workbook. |
 | officemarkup-pptx | OfficeIMO Markup | PPTX | OfficeIMO.Markup.PowerPoint | Editable | EditableEquivalent | Authors editable native typography, including family, size, color, emphasis, decoration, script, and casing, in the generated Office document; diagnostics identify profile-specific approximations and omissions. | Established | Typed profile fixtures verify parser diagnostics, editable destination artifacts, font family/size/color, bold, italic, underline variants, strike, scripts, case transforms, small caps where native, and target-specific layout. | OfficeIMO Markup is a directed authoring format rather than a lossless Office round trip; unsupported blocks and destination-only effects are diagnosed, simplified, or omitted according to exporter options. | No | `OfficeMarkupParser.Parse(markup, options).Document.ToPowerPointPresentationResult(exportOptions)` | OfficeMarkupPowerPointConversionResult | Render presentation-profile OfficeIMO Markup into an editable PowerPoint presentation. |

Docs/Compatibility/generated/package-operations.json

@@ -772,6 +772,20 @@
       "extensions":[".br",".brotli",".csv",".deflate",".gz",".gzip",".zlib"],
       "limitation":""
     },
+    {
+      "id":"conversion:dbf-csv",
+      "packageId":"OfficeIMO.CSV",
+      "formatId":"DBF/xBase",
+      "targetFormatId":"CSV",
+      "capabilityId":"dbf-csv",
+      "operation":"Convert",
+      "state":"Partial",
+      "publicApi":"DbfDataReader.Open(path), then CsvDocument.WriteDataReader(writer, reader, options)",
+      "evidence":"Independent Python-produced dBASE III, FoxPro 2 and Visual FoxPro fixtures exercise typed rows, deleted-row filtering, DBT/FPT memos, binary Base64, CSV reopen, and streaming/buffered XLSX reopen and Open XML validation.",
+      "sourceCatalog":"OfficeConversionCapabilityCatalog",
+      "extensions":[".dbf"],
+      "limitation":"Requires DBAClientX.Dbf's bounded field profiles. Native application acceptance and Visual FoxPro 31/32 generations remain unqualified. Output omits DBF indexes, deletion flags and native schema metadata; CSV nulls need a selected marker, and XLSX follows Excel numeric precision and cell text limits."
+    },
     {
       "id":"conversion:confluence-html",
       "packageId":"OfficeIMO.Confluence",
@@ -2844,6 +2858,20 @@
       "extensions":[".epub"],
       "limitation":"Image output is intentionally flattened. Unsupported source structure, remote or missing resources, and advanced source-specific layout remain approximated or diagnosed."
     },
+    {
+      "id":"conversion:dbf-xlsx",
+      "packageId":"OfficeIMO.Excel",
+      "formatId":"DBF/xBase",
+      "targetFormatId":"XLSX",
+      "capabilityId":"dbf-xlsx",
+      "operation":"Convert",
+      "state":"Partial",
+      "publicApi":"DbfDataReader.Open(path), then ExcelDocument.WriteDataReader(stream, reader, options)",
+      "evidence":"Independent Python-produced dBASE III, FoxPro 2 and Visual FoxPro fixtures exercise typed rows, deleted-row filtering, DBT/FPT memos, binary Base64, CSV reopen, and streaming/buffered XLSX reopen and Open XML validation.",
+      "sourceCatalog":"OfficeConversionCapabilityCatalog",
+      "extensions":[".dbf"],
+      "limitation":"Requires DBAClientX.Dbf's bounded field profiles. Native application acceptance and Visual FoxPro 31/32 generations remain unqualified. Output omits DBF indexes, deletion flags and native schema metadata; CSV nulls need a selected marker, and XLSX follows Excel numeric precision and cell text limits."
+    },
     {
       "id":"protection:vba-signature:excel:create",
       "packageId":"OfficeIMO.Excel",