Diff & Text Comparison Guides
Practical explanations of how diff algorithms operate, how to interpret similarity percentages, format-specific comparisons for code and documents, and how to eliminate diff noise.
Start Here — Core Fundamentals
Foundational guides on diff categories, similarity scoring, and browser privacy architecture.
- 4 min read
How to read a diff
“What am I actually looking at?”
The five categories, what a similarity score is measuring, and where the output is inference.
Read practical guideExplore → - 4 min read
Private & client-side diff
“Is it safe to compare confidential text, code, or documents online?”
Confidential text, proprietary code, GDPR compliance, and why zero server upload is the only real privacy guarantee.
Read practical guideExplore →
By File Format & Data Type
Tailored diffing techniques for source code, Word & PDF documents, JSON structures, and config files.
- 5 min read
Comparing code
“What actually changed in this refactor?”
What changed in a refactor, why renames still match, and when to distrust a high score.
Read practical guideExplore → - 4 min read
Comparing documents
“What changed between these two drafts?”
Drafts, reworded sentences, reordered sections, and dropping a .docx or PDF straight in.
Read practical guideExplore → - 4 min read
Comparing JSON
“These two JSON files should be equivalent — are they?”
Key order, reformatting, nesting, and the gap between a text diff and equivalence.
Read practical guideExplore → - 4 min read
Comparing CSV and config
“Why does my config diff say everything changed?”
Reordered keys, shifted columns, quoting, indentation, and spreadsheet vs text diffing.
Read practical guideExplore →
Refining & Cleaning Diff Results
How to diagnose moved blocks, filter whitespace noise, and handle case sensitivity accurately.
- 3 min read
Moved and reordered blocks
“I moved a section — why is it reported as deleted and re-added?”
Why relocation reads as delete-plus-add, and what order-independent matching does instead.
Read practical guideExplore → - 3 min read
Whitespace and case
“How do I stop trivial differences drowning the real ones?”
What capitalisation and indentation do to a result, and which of them there is a switch for.
Read practical guideExplore →
No guides match your search
Try clearing your search query or switching category filters.
Understanding Diff Results & Limitations
A comparison tool calculates the minimal set of additions, deletions, and moves to transform one text into another. Knowing how to interpret those signals prevents costly assumptions.
Which guide do you need?
Most people arrive at a comparison tool with one of two problems:
- •The result is confusing: A wall of marked-up rows or an unfamiliar score. Start with How to read a diff to understand the five categories and confidence markers.
- •The result has formatting noise: Reordered config keys or wrapped paragraphs. Head straight to our format guides for Documents (DOCX/PDF), JSON, or Code.
What a comparison cannot prove
A diff calculation reports purely literal and structural differences between two texts. It does not evaluate intent, correctness, or behavioral equivalence:
- •High similarity is not semantic proof: A 99% similarity score simply means very few characters were altered. Relocating a single critical variable or punctuation can entirely invert program execution or legal terms.
- •Zero differences does not guarantee identical meaning: Reformatting in white-space sensitive grammars or changing encoding can alter interpretation even when a diff reports trivial deltas.
Ready to compare your files?
Paste two texts or drag & drop Word documents (.docx), PDFs, and code directly into the comparison workbench. Everything runs 100% locally in your browser with guaranteed privacy.