Add Alt Text

add_alt_text.py

# add_alt_text.py

# import utils to load required shared libraries
from utils import inputPath, outputPath
from pdfixsdk.Pdfix import *
    
    
def AddAltText(page:PdfPage): 
    pDoc = page.GetDoc()
    content = page.GetContent()
    for i in range(content.GetNumObjects()):
        page_obj = content.GetObject(i)
        content_mark = page_obj.GetContentMark()
        is_image = type(page_obj) is PdsImage

        if is_image:
            # remove existing MCID
            page_obj.RemoveTags(kContentMarkMcid)

            # add alt text to object
            figure_dict = pDoc.CreateDictObject(False)
            figure_dict.RemoveKey('Alt')
            figure_dict.PutString("Alt", "This is some alt text")
            print("Alt text: ", figure_dict.GetText("Alt"))

            content_mark.AddTag("Figure", figure_dict, False) 
    page.SetContent() 
            
    
    

pdfix  = GetPdfix()
if pdfix is None:
    raise Exception('Pdfix Initialization fail')

doc = pdfix.OpenDoc("test.pdf", "")
if doc is None:
    raise Exception('Unable to open pdf : ' + pdfix.GetError())

# preflight document for better tagging structure (optional)
tmpl = doc.GetTemplate()
for i in range(0, doc.GetNumPages()):
    tmpl.AddPage(i, 0, None)
tmpl.Update(0, None)

# set the document title and language
title = "Supplemental Document"
lang = "en-US"

# run make accessibile
# accessibleParams = PdfAccessibleParams()
# if not doc.MakeAccessible(accessibleParams, title, lang, 0, None):
#     raise Exception(pdfix.GetError())

for i in range(doc.GetNumPages()):
    page = doc.AcquirePage(i)
    AddAltText(page)
    page.Release()
    
if not doc.Save("alt_text.pdf", kSaveFull):
  raise Exception(pdfix.GetError())

doc.Close()