Reading a result

A citation that is real but misread is still a wrong answer. Two fields on every result decide what it proves, and both are easy to skip past.

matchedIn — the main text, or an editor’s note?

Modern printings of classical works carry two layers on the page: the author’s text (the matn), and the editor’s footnotes — cross-references, gradings, and quotations from other scholars.

Your words may have been found in either.

Value	What it means
body	The words are in the author’s own text.
footnote	The words are in an editor’s footnote, not the author’s text.
both	They appear in both layers on that page.
unknown	The source did not say.

This distinction changes the claim entirely. Searching إنما الأعمال بالنيات in hadith-primary returns a hit in Sunan Abi Dawud, page 3/526 — with matchedIn: "footnote". Reading it shows why: the words are in a note quoting al-Khattabi’s commentary on the hadith. Abu Dawud did not write that sentence. Presenting it as the matn of Abu Dawud would be a fabrication built entirely out of true parts: a real book, a real page, a real quotation, and a false attribution.

When matchedIn is footnote, say so. “An editor’s note in Sunan Abi Dawud quotes al-Khattabi as saying…” is accurate. “Abu Dawud says…” is not.

relaxedToTokens — how close was the match?

Value	What was found
false	Your exact phrase, in order, as you wrote it.
true	The phrase was not found. These pages carry your words separately.

When it is true, the result also carries a note saying so in plain language:

The exact phrase was not found. These pages carry the words separately, so they answer a weaker question than the one asked.

This is the difference between “this wording is printed here” and “these words appear near each other on this page.” Both are useful; only the first supports a quotation. A page found by the looser pass may be discussing something else entirely and merely share vocabulary.

When nothing is found

If neither pass finds anything, you get an empty passages list and:

Neither the exact phrase nor its words together appear anywhere in this scope.

This is a finding, not a failure. A negative result from a scope of 8 edition-pinned hadith collections is genuine evidence — it means that wording is not in those eight books, which is often precisely what someone wants to know about a quotation circulating online.

It is not, however, evidence about the whole library. Say which scope was searched. “Not in the eight primary collections” is a real claim; “does not exist” is not one this tool can support.

Reading the excerpt

The arabic field is the passage as printed, with all markup removed. A leading … means the excerpt begins mid-sentence — it is a window onto the page, not the whole page. A trailing … means it was truncated at the length cap.

Quote it as it stands. Do not repair, re-point, or modernise the Arabic; if it looks odd, that is usually the printing, and altering it breaks the one guarantee the excerpt carries.

What a result proves, in one line

A result proves that this passage is printed in this book at this page in this edition. Everything beyond that — that it is authentic, that it is the majority view, that it answers the question asked — is a separate judgement the tool does not make and cannot make for you.