Obsidian Importer Status Tracker

[LuccaHellriegel]

NOT UNDER ACTIVE DEVELOPMENT - PROVIDED AS IS, SHOULD WORK FOR MANY CASES

Test the converter here: https://luccahellriegel.github.io/tana-import-tools/

I will use this issue to track the TODOs for the Obsidian Importer and let other people contribute. Please post all the missing features below. I will edit the issue as we go along.

If possible include an example for the found issues so that I can replicate it / more easily figure out how to integrate it.

IN PROGRESS: (holiday)

Status:

• Task Plugin Queries
• Dataview attributes without values
• footnotes: https://tanacommunity.slack.com/archives/C044X2ZC335/p1665948771836759
• normal markdown links [My Link](My%20Link.md) and also with URL (URL works out of the box)
• callouts / block quotes as own node with #block-quote
• embeds
• Dataview Attributes: https://github.com/blacksmithgu/obsidian-dataview (standard should already work, just not inline)
• Dataview Queries
• Dataview tables
• Markdown tables
• Frontmatter Comments
• Frontmatter sub-values that start with empty space instead of "-"
• improve test separation / add more separate integration tests
• proper detection of "non"-tags (obsidian has pretty clear guidlines: https://help.obsidian.md/How+to/Working+with+tags)
• perf tuning: bring down runtime for huge vault-zips from 60 sec to 30
• support for web and local version
• different file-path links (shortest: [[file]], absolute: [[folder/file]])
• tags to supertags
• aliases
• heading links
• invalid heading links (either missing or has trailing #-symbols)
• block links
• invalid block links
• codeblocks
• Frontmatter custom Attributes (need Attribute/Field Definition)
• Summary for fields
• Frontmatter tags
• Remove Frontmatter that is only Obsidian specific (publish / cssclass / aliases)
• daily notes / date notes
• todos
• plain markdown image links with urls

Backburner:

• images (need Firebase URLs, not possible without an API - possibly your own file-upload API)
• Headings with Heading Links or normal links in them
• relative file paths: [[./file]] (this would require checking the folder again and saving filepath<->UID pairs)
• callouts / block quotes - while they are split into their own node, they are not native to Tana and will be marked with #block-quote
• #[[bla bla]] links -> currently we just remove the link here

[TanaRob]

Aliases

[[page name|alias]]

Header references

[vicrdguez]

Have obsidian callouts an equivalent in Tana? or they will be translated just as a quote block?

[LuccaHellriegel]

• Added some documentation about how the importer works (much more complicated conversion than another graph format)
• Folder Structure is taken into account (nodes are nested under their respective folder-nodes)
• Links inside files are replaced with appropriate UIDs (this means, same folder/file names work now)
• Missing nodes are created separately (up for debate, but they are tracked - easy to change now)

[tomasztatar]

Hi.
I had problems with importing TIF files when name attributes contained full Windows paths for example:
"name": "C:\Users\qwerty\OneDrive - company\Documents\iCloud-Obsidian-notes"
or
"name": "C:\Users\qwerty\OneDrive - company\Documents\iCloud-Obsidian-notes\--- PARA",

Tana refused to import it. When I manually replaced names to:
"name": "iCloud-Obsidian-notes",
"name": "--- PARA",
it worked.

I am not sure what requirements does Tana enforce on name attribute.

As for web zip converter, I always have error:

An Error occured.

It probably has to do with a not properly zipped vault.
Please check the text at the top again.

Content of Error (please post in the Slack and tag me):

TypeError: Cannot read properties of undefined (reading 'children')
at lu.readDirectory (https://luccahellriegel.github.io/tana-import-tools/assets/index.b5f269bd.js:2:37178)
at wp (https://luccahellriegel.github.io/tana-import-tools/assets/index.b5f269bd.js:25:1465)
at cc (https://luccahellriegel.github.io/tana-import-tools/assets/index.b5f269bd.js:25:1794)
at bp (https://luccahellriegel.github.io/tana-import-tools/assets/index.b5f269bd.js:36:10)
at async HTMLInputElement.<anonymous> (https://luccahellriegel.github.io/tana-import-tools/assets/index.b5f269bd.js:44:687)

I am willing to help with troubleshooting it.

[LuccaHellriegel]

Hi @tomasztatar

as for the second problem, did you zip the full vault folder and keep the zip name the same like described?

[tomasztatar]

Hi @LuccaHellriegel

No, I missed requirement about naming zip file. Now I retested and web converter worked fine. Additionally TIF file does not contain full Windows paths and imports successfully into Tana. Thanks.

But I have another issue - my Obsidian files contain tags, usually at the end, in separate paragraph. They convert to Tana tags, but are located as separate sub-node. That means that listing tagged nodes list their empty tag-only nodes. I would expect that file-level node is tagged (as it is in Obsidian).

[SutanuNandigrami]

#29

[LuccaHellriegel]

Not under active development anymore.