Preeti to Unicode Conversion Errors

The source is not actually Preeti

Kantipur, PCS Nepali, Sagarmatha and other legacy Nepali fonts can look similar while using different internal mappings. Running those characters through a Preeti mapping produces the wrong Nepali letters.

Select a suspicious word in the original document and inspect its font. If multiple fonts appear inside one paragraph, separate the runs by font before conversion.

Vowel signs appear in the wrong place

The short-i sign and some visual marks are stored differently in legacy fonts. A partial copy, an unsupported sequence or a manual edit to raw Preeti codes can leave a matra attached to the wrong consonant.

Return to the original source and convert the complete word or sentence. Avoid converting fragments cut through a combined sequence. Then compare the Unicode result with the visible Preeti original.

Reph, half letters and conjuncts are broken

• Check words containing र् and common conjuncts such as क्ष, त्र and ज्ञ.
• Convert complete words so the mapper has the surrounding character sequence.
• Confirm that copy and paste did not introduce spaces between legacy codes.
• Correct rare forms directly in Unicode after comparing with the source.

English letters remain after conversion

Some English content may be intentional, especially email addresses, URLs, abbreviations and mixed-language headings. Other leftover characters can indicate an unmapped glyph or a different source font.

Review leftover Latin characters instead of deleting them automatically. Compare their position with the source and preserve legitimate English terms.

Punctuation and spacing look incorrect

Old documents often use manual spaces to imitate Devanagari placement. Those spaces may become unnecessary after Unicode shaping. Quotation marks, danda marks, numerals and tab characters can also differ between applications.

Normalize repeated spaces only after conversion, preserve paragraph breaks, and review punctuation visually in the final destination such as Word, a website CMS or a PDF.

A safe correction workflow

• Keep the original file unchanged.
• Convert one representative paragraph first.
• Compare source and result side by side.
• Correct the Unicode output rather than altering raw Preeti codes.
• Record recurring corrections when converting a large archive.
• Have important documents reviewed by a second person.