# Status Broadcast System (sbs)
This is only a test.

`sbs` looks at data from Jira, formats it, combines it with a prompt which can be fed into
an LLM for creating status.


## Usage
### Top Level
```
usage: sbs [-h] {jql,csv} ...

Turns Jira output into a Gemini prompt to generate status

positional arguments:
  {jql,csv}   commands
    jql       API call to Jira to pull data
    csv       Read an exported CSV file from Jira

options:
  -h, --help  show this help message and exit
```

### jql
`jql` is what most people will be looking for. It takes a JQL query and Jira API endpoint and pulls the data.

```
usage: sbs jql [-h] [-q QUERY] [-e ENDPOINT] [-t TOKEN] [-p PROMPT] [-i] mapping output

positional arguments:
  mapping               Mapping file for Jira Account=Name
  output                The file to write

options:
  -h, --help            show this help message and exit
  -q, --query QUERY     Text or @file of the jql query to use
  -e, --endpoint ENDPOINT
                        Jira API endpoint
  -t, --token TOKEN     Access token for the Jira API
  -p, --prompt PROMPT   Text or @file prompt to use
  -i, --ignore-unknowns Ignores issues that don't have a known assignee
```

### csv
`csv` is a less used command used for those who don't have API access to their Jira instance. It looks at an exported CSV file from Jira as it's data.
```
usage: sbs csv [-h] [-p PROMPT] [-i] csvfile mapping output

positional arguments:
  csvfile               Path to the CSV file to read
  mapping               Mapping file for Jira Account=Name
  output                The file to write

options:
  -h, --help            show this help message and exit
  -p, --prompt PROMPT   Text or @file prompt to use
  -i, --ignore-unknowns Ignores issues that don't have a known assignee
```

## Example

### JQL From Source
We will ignore unassigned tickets, load a prompt from the file myprompt.txt, load a Jira query from the file myquery.txt, use mappingfile.txt as our mapping file and output the results to outputfile.txt.

Since we are using the jql datasource we'll provide it the API endpoint and our access token.
```
$ python3 sbs/cli.py jql \
  -i \
  -p @myprompt.txt \
  -q @myquery.txt \
  mappingfile.txt \
  outputfile.txt \
  -e https://issues.redhat.com/rest/api/2 \
  -t <TOKEN>
```

### CVS From Source
We'll give it a prompt on the command line, use mappingfile.txt as our mapping file and output the results to outputfile.txt.

Since we are using the csv datasource we'll provide it the name of the file we exported from Jira.
```
$ python3 sbs/cli.py csv \
  -p "Summarize the following as markdown. Limit each bullet point to a maximum of two sentences."
  jira_export.csv \
  mappingfile.txt \
  outputfile.txt
```

## mapping
The mapping file is a simple key=value file seperated by newlines which maps a Jira username to a human name. It's used to make the output used in the prompt more natural for human reading.

**Note**: malformed lines are ignored and the farthest left `=` is used as the seperator.

### Format
```
key=value\n
key=value\n
key=value\n
...
```

### Example mapping File
```
$ cat mapping_file.txt
hyamamo=Himari Yamamoto
cosmic_coder=Ricardo Ruiz
metalHEAD42@example.com=Bo Smith
$
```