How to edit Hindi (Devanagari) text in a PDF on your phone
Trying to edit a Hindi PDF often results in broken words. Matras (vowel signs) detach from their consonants, half-letters (conjuncts) expand into full letters with halants, and the document looks entirely unprofessional.
The challenge with Devanagari in PDFs
Devanagari is a complex script. Characters combine to form ligatures, and vowel markers can appear before, after, above, or below the base consonant. Most standard PDF editors lack the text-shaping engine required to handle these rules. When you add text, they just place the raw characters side-by-side.
The proper way to edit Hindi PDFs
Editrix PDF includes native support for Devanagari text shaping. When you edit a Hindi word, the app correctly combines consonants and matras, preserving the exact visual appearance of the language.
Here is how to do it:
- Open the PDF in the Editrix app.
- Tap the text block you wish to modify.
- Make your edits using your standard Hindi keyboard.
- The text will shape correctly as you type. Tap save.
Why Devanagari is harder than it looks
The difficulty is that what you type and what gets drawn are not in the same order, and there are not the same number of them.
Take कि. You type the consonant क first and the vowel ि second, but the vowel is drawn to the left of the consonant. The renderer has to reorder them. Conjuncts go further: क् + ष becomes the single ligature क्ष, where two characters produce one glyph that resembles neither. A PDF stores only the final glyph and its position, so reversing that to recover "क्ष" - let alone editing it - needs the font's own rules, not a character table.
Editrix shapes Devanagari through HarfBuzz, the shaping engine Android itself uses, so reordering, conjunct formation, matra placement and the i-matra's leftward jump all follow the embedded font's OpenType rules rather than a guess. The glyph ids HarfBuzz produces are written straight back into the PDF.
Common failures and what they mean
- Matras float away from their consonant. No shaping was applied; the characters were placed side by side in typing order.
- Conjuncts break into full letters with visible halants (क् ष instead of क्ष). The ligature substitution rules in the font were not applied.
- Boxes or missing characters. The embedded font is subset - PDF creators strip out unused glyphs - so a conjunct the original document never contained simply is not in the file.
OCR for scanned Hindi documents
If your PDF is a scan or photograph, there is no text to edit, only an image. Editrix includes OCR with a Devanagari model, so you can convert the page to selectable, searchable text first. Expect to proofread: Devanagari OCR is markedly less accurate than Latin, particularly on conjuncts and on low-contrast scans.
When another tool is better
If you need to translate an entire document or rewrite pages of text, a word processor is the right tool - convert to Word, edit on a desktop, export back. Editrix is the better choice when the layout must survive: correcting a name on a certificate, fixing a date on an official letter, or filling a government form on your phone without the formatting collapsing.