ObsidianLoader doesn't Parse Frontmatter of Obsidian Files according to YAML, but line-by-line and only for key-value pairs

[effi]

System Info

Python 3, Google Colab/Mac-OS Conda Env, Langchain v0.0.173, Dev2049/obsidian patch #4204

Who can help?

No response

Information

• [ ] The official example notebooks/scripts
• [X] My own modified scripts

Related Components

• [ ] LLMs/Chat Models
• [ ] Embedding Models
• [ ] Prompts / Prompt Templates / Prompt Selectors
• [ ] Output Parsers
• [X] Document Loaders
• [ ] Vector Stores / Retrievers
• [ ] Memory
• [ ] Agents / Agent Executors
• [ ] Tools / Toolkits
• [ ] Chains
• [ ] Callbacks/Tracing
• [ ] Async

Reproduction

See Colab: https://colab.research.google.com/drive/1YSfKTQ92RZEJ-z-hlwHL7f5rQ-evTgx4?usp=sharing

1. Have Obsidian Files with Front_Matter other then simple key_values
2. Import ObsidianLoader and create loader with the Obsidian Vault
3. load documents

YAML frontmatter is parsed to: metadata={'source': 'Demo-Note.md', 'path': 'Obsidian_Vault/Demo-Note.md', 'created': 1684513990.4561925, 'last_modified': 1684513964.4224126, 'last_accessed': 1684514007.4610083, 'normal': 'frontmatter is just key-value based', 'list_frontmatter': '', 'dictionary_frontmatter': '', 'can': 'contain', 'different': 'key-value-paris', 'and': '[even, nest, these, types]'}

Expected behavior

YAML Frontmatter should be parsed according to YAML rules

normal: frontmatter is just key-value based list_frontmatter:

• can
• be
• an array dictionary_frontmatter: can: contain different: key-value-paris and: [even, nest, these, types]

---

-->

{ 
    'norma'l: 'frontmatter is just key-value based,
    'list_frontmatter': ['can', 'be', 'an array'],
    'dictionary_frontmatter' : {
        'can' : 'contain',
        'different': 'key-value-pairs',
        'and' : ['even', 'nest', 'theses', 'types]
    }
}