Avro to Bijoy

What Avro writes to the clipboard

• Type or paste a line in Avro and copy it.
• If the text is standard Unicode Bengali, this page is the right tool and there is no extra step to skip.
• If it converts to Latin letters on the output side, you are holding legacy ANSI text — and that came from somewhere else, not from Avro.

Why people think Avro to Bijoy needs a special converter

Two things feed the confusion. Older builds of Avro could also write legacy ANSI for printing, so "Avro" and "Bijoy" overlap in memory for anyone who used them. Separately, the phrase "Bijoy layout" gets used loosely for "Bengali text as it appears in Bijoy", which is not an input format at all.

Whether Ridmik, Gboard and Avro produce identical Bengali text

Yes. Ridmik, Gboard, Google Input Tools and Windows' own Bengali layout all emit the same Unicode code points for the same letters. They differ in how you type, not in what gets stored, so one paste-and-convert step serves all of them.

Setting the font in Microsoft Word

• Paste the converted text into Word.
• Select it and set the font to SutonnyMJ — or to whichever MJ font your press specified.
• Check both the Latin and the complex-script font slots for that selection. Word keeps them separate, and filling in only one leaves part of the text unmapped.
• Save, then reopen the file before sending it out. If it still renders correctly after reopening, the assignment really did stick.

Setting the font in InDesign and Illustrator

• InDesign: define a character style carrying the MJ font once, then apply that style to the converted text. A character style is what keeps a long document consistent, and it survives later edits.
• InDesign: if the text looks wrong on the page but fine in Story Editor, the font is coming from a paragraph style with higher precedence.
• Illustrator: select the text with the Type tool and set the font in the character panel. Illustrator will not substitute quietly, so a missing font is reported rather than hidden.

When the font cannot be installed

Three routes, in order of how much you lose. Ask the recipient to install the MJ font — for a publisher or a press this is normally trivial, and it is the only route that keeps the file editable. Or match the font the rest of the document already uses. Or convert the Bengali text to outlines before making the PDF: that always renders, but the text stops being selectable and searchable, so it is a last resort for a final proof rather than a working file.

Bengali text in Google Docs and online editors

Online editors cannot be relied on for this. They have no access to an installed legacy font, so they fall back to a Unicode font and reorder the glyphs, and the text ends up looking wrong. That is not the conversion failing; it is an editor that cannot render the output correctly. Use it for editing, then set the font in desktop software before final output.

Reading the "needs font" badge on the output

If the badge appears, the output contains real ANSI code points and does need an MJ font to render. If it never appears and the output still shows Bengali script rather than Latin letters, then no conversion happened — which almost always means there was nothing to convert.

What the unmapped-character warning is telling you

Some code points have no legacy equivalent and are passed through untouched. Any that appear are listed, and they will render as wrong or missing glyphs in an MJ font — so read the list rather than skimming past it.

Checking reph and the i-kar signs by eye

• A reph (র্) should sit above and after its consonant cluster, not in front of it.
• The signs ি and ে belong before the consonant they attach to, including inside conjuncts.
• Marks follow the cluster, never the whole word — a mark drifting to a word boundary is the classic sign of a broken paste.

Telling Unicode Bengali from mangled legacy text

• Unicode Bengali shows as Bengali letters in the browser, in any app, with no font installed.
• Legacy text shows as Latin letters and only becomes Bengali once an MJ font is applied.
• If you expected Bengali letters and got Latin ones, you are holding legacy text — the reverse direction is the page you want.

Recovering text that came back from a legacy workflow

Legacy text that arrived from an old workflow can be read back into Unicode on the Bijoy to Unicode page. That direction accepts the wider set of variants produced by different MJ family fonts and keyboard builds, so text from an older template still comes back correctly.

Mixed documents where only part of the text is legacy

Mixed documents are the awkward case, and they are common in imported files. Do not paste the whole thing and convert at once — the Unicode half would be treated as legacy and corrupted. Convert the two halves separately, or copy the mangled parts out, recover them, and paste them back.

There is no official Bijoy code chart

The encoding was never published as a standard. It maps Bengali glyphs onto Latin code points, and different tools settled on different bytes for the rare conjuncts, because there was no authority to settle them the same way.

Which bytes this converter emits, and which it still reads

This converter emits the mainstream variant, the one the most widely installed MJ fonts render correctly. It also reads the variants other tools produce, so text coming from an older tool is still understood. That combination is why the comparison can be made at all — it comes from matching the font rather than copying another site's table.

University and thesis submissions

Departments in Bangladesh commonly ask for a specific MJ font, and the rejection reason is usually that the text is machine-readable Unicode rather than the legacy form. Convert first, set the font, then confirm the requirement is about the font and not about the text.

Government forms and official documents

The same pattern applies with less room to negotiate. The form template is fixed, so the font has to match whatever it uses. If a form will not accept the text, check the font before assuming the conversion was wrong.

Books, magazines and commercial printing

Print work goes through layout software rather than Word, so it is worth converting before the file enters that stage and setting up a character style as described above. Never send a PDF with the text still in Unicode Bengali and expect the typesetter to fix it.

What happens to tables, images and styles

Only the text is rewritten. Tables, images, styles, headers, footers, comments, tracked changes and formulas are left exactly as they were, and you download the same document back. Merged text runs are reported rather than applied silently, because a formatting boundary falling inside a word can change the output.

Editing the converted document afterwards

After conversion the text is legacy ANSI, so the practical rules change. Type new Bengali through the converter rather than straight into the document, keep the font assignment when applying styles, and if a large edit is needed it is usually safer to convert the original again from the source text.

Some conjuncts come out as boxes

• The font is not applied to that run — check the complex-script slot as well, not just the Latin one.
• The font is an old build that does not contain those conjuncts. Try the font the press specified.
• The text uses a rare form that no MJ font covers. Compare against a sample your publisher already accepts.

The output looks right but the press rejects it

The screen and the press are judging different things. What a press needs is the legacy code points plus a font matching its own template, and a correct-looking result on screen can still be Unicode Bengali. Compare the actual document with a file the press has accepted before, rather than comparing how it looks.

Text converts correctly but search or spellcheck breaks

This is expected rather than a fault. Legacy ANSI is not standard Bengali Unicode, so spellcheck, find-and-replace and site search treat it as Latin text, and screen readers read it as nonsense. Keep the Unicode source as the working copy and convert only the version that has to be handed over.