Blog Post Migration from Blogger

Re-transcribe poorly migrated blog posts from the original Blogger source at martinondotnet.blogspot.com into clean markdown.

Overview

The blog was originally on Blogger, migrated to Hugo (losing code blocks, mangling formatting), then to Astro. Posts with source: hugo need re-transcription from the original Blogger source. Posts with source: blogger are already done.

Known Issues in Hugo-Sourced Posts

•Empty code blocks - Code lost during Blogger-to-Hugo migration (just whitespace between fences)
•Old Hugo frontmatter embedded in the content body below the Astro frontmatter
•Mangled lists - 5-space indent + * instead of standard - or 1.
•Missing descriptions - description: ''
•Wrong originalUrl - Points to codifice.dev instead of blogspot
•Unresolved shortened URLs - bit.ly, goo.gl, t.co links still present
•HTML artifacts - &, stray <a> tags, encoded entities

Sitemap

All source posts: https://martinondotnet.blogspot.com/sitemap.xml (62 URLs)

Slug Matching

Blogspot URL path slug matches the repo filename slug after the date prefix:

code

Blogspot: /2010/03/aspnet-and-custom-error-pages-seo.html
Repo:     2010-03-20-aspnet-and-custom-error-pages-seo.md

Migration Workflow Per Post

1. Read the Existing Repo File

bash

# Identify issues: empty code blocks, mangled formatting, embedded Hugo frontmatter
Read src/content/blog/{slug}.md

Note what content exists, what's broken, and what images are already referenced.

2. Fetch the Blogger Source

bash

WebFetch https://martinondotnet.blogspot.com/{year}/{month}/{slug}.html

Request the FULL content including all code blocks, images, and links. The Blogger markup is complex with nested <div>, <span> with inline styles, and Blogger-specific classes. Do not attempt to regex/script the conversion - manually transcribe to clean markdown.

3. Transcribe to Clean Markdown

Rules:

•Transcribe content faithfully from the Blogger source
•Correct spelling errors only (not grammar or style)
•Use proper markdown: # headings, - lists, ``` fenced code blocks with language tags
•Code blocks MUST include the language identifier (e.g., ```csharp, ```xml, ```sql)
•Preserve the author's voice, emphasis (**bold**, _italic_), and structure
•Remove any embedded old Hugo frontmatter from the content body
•Convert HTML entities to actual characters

4. Resolve Shortened Links

For every bit.ly, goo.gl, t.co, or other shortened URL:

bash

# Follow redirects to get the final destination
curl -sI -L "https://bit.ly/xxxxx" 2>&1 | grep -i "^location:"

Replace with the final destination URL. If the destination is dead, keep the original with a comment: [link text](https://original-url "Link may be broken").

5. Migrate Images

Inline images from Blogger CDN (e.g., blogger.googleusercontent.com, bp.blogspot.com):

bash

# Download to src/assets/blog/ with slug-prefixed descriptive name
curl -sL "https://blogger.googleusercontent.com/img/..." \
  -o "src/assets/blog/{slug}-{descriptive-name}.png"

Naming convention: {post-slug}-{descriptive-name}.{ext}

Example: diagnosing-tricky-aspnet-production-image_thumb.png

Reference in markdown:

markdown

![Description](../../assets/blog/{slug}-{descriptive-name}.png)

Check if images already exist in src/assets/blog/ before downloading - many were migrated during the Hugo phase.

6. Migrate Downloadable Assets

Check the OneDrive folder for attachments referenced in the post:

bash

ls "/mnt/c/Users/mcj/OneDrive/Public/MartinOnDotNet/"

Available files:

•AdvancedControlAdapter.zip
•CodeBehindRemover.zip
•IIS7Web.zip
•MasterPageBodyClass.zip
•QLXFilters.zip
•TypeConversion.zip
•Verification Tests.zip

Copy matched assets to public/downloads/{slug}/:

bash

mkdir -p public/downloads/{slug}
cp "/mnt/c/Users/mcj/OneDrive/Public/MartinOnDotNet/{file}" public/downloads/{slug}/

Reference in markdown:

markdown

[Download the source code](/downloads/{slug}/{file})

7. Update Frontmatter

yaml

---
title: 'Exact title from Blogger source'
description: 'Write a meaningful 1-2 sentence description from the post content'
pubDate: YYYY-MM-DD # Keep existing date
updatedDate: YYYY-MM-DD # Keep if exists, add from Blogger if visible
tags: ['tag1', 'tag2'] # Keep existing or improve from Blogger labels
source: blogger # Change from 'hugo' to 'blogger'
originalUrl: 'http://martinondotnet.blogspot.com/{year}/{month}/{slug}.html'
heroImage: ../../assets/blog/hero-images/{slug}.jpg # Keep existing
---

Key changes:

•source: hugo -> blogger
•originalUrl: codifice.dev URL -> blogspot URL
•description: Fill in if empty
•title: Verify matches Blogger source (may have been truncated)

8. Review the Migration

After writing the file, verify:

code

- [ ] Content matches Blogger source faithfully
- [ ] All code blocks have content (no empty blocks) and language identifiers
- [ ] All images are present in src/assets/blog/ and render correctly
- [ ] All links are functional (no shortened URLs remain)
- [ ] No HTML artifacts or entities remain
- [ ] No embedded Hugo frontmatter in content body
- [ ] Frontmatter is complete and correct
- [ ] Description field is filled in
- [ ] source is 'blogger' and originalUrl points to blogspot
- [ ] Markdown passes lint: npm run lint:md
- [ ] Post builds without errors: npm run build

Quick Reference: Post Status

Status	Count	Criteria
Done (blogger)	4	`source: blogger` - already re-transcribed
Needs migration	57	`source: hugo` with matching blogspot URL
No repo file	1	`martin-on-net-is-changing-to` (migration announcement - skip)

To find posts still needing migration:

bash

grep -l "source: hugo" src/content/blog/*.md

Common Pitfalls

•WebFetch may summarize - If code blocks are truncated, request specific sections
•Blogger code uses SyntaxHighlighter - Look for <pre class="brush:"> blocks
•Images may be thumbnails - Blogger often wraps thumbnails in links to full-size images; use the full-size URL
•Download links may point to dead services - Check OneDrive folder as fallback
•Some posts reference other posts - Update internal links to the new Astro blog URL format: /blog/{slug}