Skip to content

Alignment of word bounding boxes #4

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
chryslovelace wants to merge 4 commits into
base: main
Choose a base branch
from
Open

Alignment of word bounding boxes #4

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
00dc060
font size calculation, and moved xml file saving up out of processor
chryslovelace Sep 8, 2025
6412a84
try swapping x coords
chryslovelace Sep 19, 2025
803b9dd
try adjusting bboxes based on char length
chryslovelace Sep 23, 2025
d283ce0
version number increase
chryslovelace Sep 25, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 2 additions & 10 deletions HocrHeaderFixer.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -19,20 +19,12 @@ public void Init()
Console.WriteLine("Fixing hOCR file headers...");
}

public void Process(string hocrFile, XDocument hocrXml)
public bool Process(string hocrFile, XDocument hocrXml)
{
XNamespace ns = "http://www.w3.org/1999/xhtml";
var head = hocrXml.Element(ns + "html").Element(ns + "head");
head.Element(ns + "title").Value = "Image: " + FileName ?? Regex.Replace(Path.GetFileName(hocrFile), "_HOCR.shtml$", "_JP2.jpg");
head.Add(new XElement(ns + "meta", new XAttribute("name", "ocr-system"), new XAttribute("content", HtrId.HasValue ? $"Transkribus-HtrId:{HtrId}" : "Transkribus")));
using var fileStream = File.Open(hocrFile, FileMode.Truncate);
var writer = XmlWriter.Create(fileStream, new XmlWriterSettings
{
// need to specify false here to stop it from emitting a byte order mark
Encoding = new UTF8Encoding(false),
Indent = true
});
hocrXml.Save(writer);
writer.Close();
return true;
}
}
12 changes: 11 additions & 1 deletion IHocrXmlProcessor.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,6 +2,16 @@

interface IHocrXmlProcessor
{
/// <summary>
/// Used to set up any initial state, such as creating directories
/// </summary>
void Init();
void Process(string hocrFile, XDocument hocrXml);

/// <summary>
/// Receives each hOCR xml tree for processing
/// </summary>
/// <param name="hocrFile">Path to the hOCR file</param>
/// <param name="hocrXml">Xml tree representing the hOCR file</param>
/// <returns>True if the xml tree was modified and needs to be saved back to the file.</returns>
bool Process(string hocrFile, XDocument hocrXml);
}
7 changes: 4 additions & 3 deletions OcrGenerator.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,14 +16,14 @@ public void Init()
Directory.CreateDirectory(OcrDirectory);
}

public void Process(string hocrFile, XDocument hocrXml)
public bool Process(string hocrFile, XDocument hocrXml)
{
var text = new StringBuilder();
XNamespace ns = "http://www.w3.org/1999/xhtml";
foreach (var paragraph in hocrXml.Descendants(ns + "p"))
{
var lines = paragraph.Elements(ns + "span").Where(span => span.Attribute("class")?.Value == "ocr_line");
foreach (var line in lines)
var lines = paragraph.Elements(ns + "span").Where(span => span.Attribute("class")?.Value == "ocr_line");
foreach (var line in lines)
{
var words = line.Elements(ns + "span").Select(span => span.Value);
text.AppendJoin(' ', words);
Expand All @@ -33,5 +33,6 @@ public void Process(string hocrFile, XDocument hocrXml)
}
var ocrFile = Path.Join(OcrDirectory, Regex.Replace(Path.GetFileName(hocrFile), "_HOCR.shtml$", "_OCR.asc"));
File.WriteAllText(ocrFile, text.ToString().Trim());
return false;
}
}
28 changes: 22 additions & 6 deletions Processor.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,6 +2,7 @@
using System.Runtime.InteropServices;
using System.Text;
using System.Text.RegularExpressions;
using System.Xml;
using System.Xml.Linq;
using Flurl.Http;

Expand Down Expand Up @@ -50,10 +51,12 @@ public async Task<byte[]> ProcessSinglePage(Uri fileUri, MicroservicePageOptions
Directory.CreateDirectory(Jp2Directory);
await File.WriteAllBytesAsync(Path.Join(Jp2Directory, Path.GetFileName(fileUri.LocalPath)), sourceFile);
await ConvertJp2sToJpgs();
var page = await SendSinglePageToTranskribus(options);
var page = await SendSinglePageToTranskribus(options);
await GetSinglePageTranskribusAltoXml(page);
await ConvertAltoToHocr();
ProcessHocrXml(new HocrHeaderFixer(options.HtrId, Path.GetFileName(fileUri.LocalPath)));
ProcessHocrXml(
new HocrHeaderFixer(options.HtrId, Path.GetFileName(fileUri.LocalPath)),
new WordAlignmentFixer());
var hocrFile = Directory.EnumerateFiles(HocrDirectory).Single();
return await File.ReadAllBytesAsync(hocrFile);
}
Expand Down Expand Up @@ -110,14 +113,14 @@ public async Task CheckProgress(IdCrudOptions options)
DeleteDirectoryIfExists(OcrDirectory);
}
}

public async Task CreateOcrDatastreamsFromHocr(OcrOptions options)
{
string pidFilePath = null;
try
{
pidFilePath = options.PidFile is null
? await GetPagePids(options, options.Pid)
pidFilePath = options.PidFile is null
? await GetPagePids(options, options.Pid)
: Path.GetFullPath(options.PidFile);
await GetHocrDatastreams(options, pidFilePath);
ProcessHocrXml(new OcrGenerator(OcrDirectory));
Expand Down Expand Up @@ -401,9 +404,22 @@ void ProcessHocrXml(params IHocrXmlProcessor[] processors)
foreach (var hocrFile in Directory.EnumerateFiles(HocrDirectory))
{
var xml = XDocument.Load(hocrFile);
var xmlModified = false;
foreach (var processor in processors)
{
processor.Process(hocrFile, xml);
xmlModified |= processor.Process(hocrFile, xml);
}
if (xmlModified)
{
using var fileStream = File.Open(hocrFile, FileMode.Create);
var writer = XmlWriter.Create(fileStream, new XmlWriterSettings
{
// need to specify false here to stop it from emitting a byte order mark
Encoding = new UTF8Encoding(false),
Indent = true
});
xml.Save(writer);
writer.Close();
}
}
}
Expand Down
46 changes: 46 additions & 0 deletions WordAlignmentFixer.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,46 @@
using System.Text.RegularExpressions;
using System.Xml.Linq;

partial class WordAlignmentFixer : IHocrXmlProcessor
{
public void Init()
{
Console.WriteLine("Fixing horizontal word alignment...");
}

public bool Process(string hocrFile, XDocument hocrXml)
{
XNamespace ns = "http://www.w3.org/1999/xhtml";
foreach (var paragraph in hocrXml.Descendants(ns + "p"))
{
var lines = paragraph.Elements(ns + "span").Where(span => span.Attribute("class")?.Value == "ocr_line");
foreach (var line in lines)
{
var lineAttributes = line.Attribute("title").Value;
var lineBbox = BboxRegex().Match(lineAttributes).Groups
.Cast<Group>()
.Skip(1)
.Select(g => int.Parse(g.Value))
.ToArray();
var words = line.Elements(ns + "span");
// # of characters = sum of word lengths + the spaces between
var lineCharLength = words.Sum(w => w.Value.Length) + words.Count() - 1;
var linePixLength = lineBbox[2] - lineBbox[0];
var pixPerChar = linePixLength / lineCharLength;
var currentPosition = lineBbox[0];
foreach (var word in words)
{
var newLeft = currentPosition;
var newRight = currentPosition += word.Value.Length * pixPerChar;
var newBbox = $"bbox {newLeft} {lineBbox[1]} {newRight} {lineBbox[3]}";
word.Attribute("title").SetValue(BboxRegex().Replace(word.Attribute("title").Value, newBbox));
currentPosition += pixPerChar;
}
}
}
return true;
}

[GeneratedRegex(@"bbox (\d+) (\d+) (\d+) (\d+)")]
private static partial Regex BboxRegex();
}
2 changes: 1 addition & 1 deletion transkribus-process.csproj
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,7 @@
<TargetFramework>net8.0</TargetFramework>
<RootNamespace>transkribus_process</RootNamespace>
<ImplicitUsings>enable</ImplicitUsings>
<Version>1.1.0</Version>
<Version>1.2.0</Version>
</PropertyGroup>

<ItemGroup>
Expand Down