simonw
Clicking 'Run SQL' from extneded query.html page results in URL not found error
Version: 1.0a19
I see the following initial query form button, (the point to note here is that we're not in the query.html extension yet.)
on the database's home page http://192.168.0.203/sam_datasette/rm_toucans_slice
The initial query page's form looks like this
<html>
<body>
<!--StartFragment-->
<form class="sql core" action="/sam_datasette/rm_toucans_slice/-/query" method="get">
--
| <h3>Custom SQL query</h3>
| <p><textarea id="sql-editor" name="sql">select * from rm_rnb_history_pres</textarea></p>
| <p>
| <button id="sql-format" type="button" hidden>Format SQL</button>
| <input type="submit" value="Run SQL">
| </p>
| </form>
<!--EndFragment-->
</body>
</html>
After running the query, the form looks like
The query page form after the query has been run and landed in an extended query.html looks like this
<form class="sql core" action="/sam_datasette/rm_toucans_slice" method="get">
--
| <h3>Custom SQL query returning 50 rows
| <span class="show-hide-sql">(<a href="/sam_datasette/rm_toucans_slice/-/query?sql=++++++select%0D%0A++rowid%2C%0D%0A++id%2C%0D%0A++tx_lng%2C%0D%0A++tx_lat%2C%0D%0A++rx_lng%2C%0D%0A++rx_lat%2C%0D%0A++timestamp%2C%0D%0A++strftime%28%27%25Y%25m%25d%27%2C+timestamp%29+as+date%2C%0D%0A++strftime%28%27%25H%25M%27%2C+timestamp%29+as+time%2C%0D%0A++midpoint_lat%28tx_lat%2C+tx_lng%2C+rx_lat%2C+rx_lng%29+as+ap_lat%2C%0D%0A++midpoint_lng%28tx_lat%2C+tx_lng%2C+rx_lat%2C+rx_lng%29+as+ap_lng%2C++%0D%0A++haversine%28tx_lat%2C+tx_lng%2C+rx_lat%2C+rx_lng%29+as+length%2C%0D%0A++dB%2C%0D%0A++frequency%2C%0D%0A++Spotter%2C%0D%0A++Country%2C%0D%0A++State%2C%0D%0A++County%2C%0D%0A++City%2C%0D%0A++QSL_Sent%2C%0D%0A++QSL_Rx%2C%0D%0A++QSL_link%2C%0D%0A++QSL_rx_link%2C%0D%0A++tx_rst%2C%0D%0A++park%2C%0D%0A++p2p_s2%2C%0D%0A++call%2C%0D%0A++skcc%2C%0D%0A++naqcc%2C%0D%0A++f2m%2C%0D%0A++launchangle%2C%0D%0A++%2722%27+as+elev_tx%0D%0Afrom%0D%0A++rm_rnb_history_pres%0D%0Awhere+%0D%0A++dB+%3E+100%0D%0A++and+timestamp+%3E+%272025-04-26T00%3A25%3A00%27%0D%0A++and+timestamp+%3C+%272025-04-27T01%3A55%3A00%27%0D%0Aorder+by%0D%0A++timestamp%0D%0A&_hide_sql=1">hide</a>)</span>
| </h3>
|
|
|
| <p><textarea id="sql-editor" name="sql" style="height: 44em"
| > select
| rowid,
| id,
| tx_lng,
|
Notice that the form action has changed and no longer contains the breadcrum. When the 'Run SQL' button is clicked again with no other changes, the page isn't found
Also notice that the not found address:
http://192.168.0.203/rm_toucans_slice/-/query?sql=+
no longer contains sam_datasette, while the previous address did:
http://192.168.0.203/sam_datasette/rm_toucans_slice/-/query?sql=+++
sam_datasette being stripped is causing the URL not found error
The first question: Is this caused by the template that extends query.html. I'm moving it to the side by renaming it query_mytest.html.
I am no longer using the customized query.html, but still getting the same error. The normal queqry page looks like
While the customized page looks like
The same error occurs for both pages.
The next question is if this is caused by #2454 I'm rolling my install back to 1.0a16
That did not fix the issue. Still seeing the same error and cause.
Changing to localhost does fix the issue
http://127.0.0.1:8001/rm_toucans_slice/-/query?sql=++++++select%0D%0A+
but means I can't run with other boxes on the network.
The command line is
python -m datasette glotec_slice.db rm_toucans_slice.db --plugins-dir=plugins --template-dir plugins/templates --root --static assets:static-files/ --setting sql_time_limit_ms 1700000 --setting max_returned_rows 100000 --cors --crossdb --setting base_url /sam_datasette/ -p 8001
Note: both versions of ProxyPass mentioned below resulted in the same issue above on the 1.0+ versions tried above.
This might be a local issue. On version 0.65.1, (where the above issue is not happening), when I configure httpd.conf as per the instructions:
ServerName 192.168.0.203:80
ProxyPass /sam_datasette/ http://127.0.0.1:8001/sam_datasette/ connectiontimeout=300 timeout=300
ProxyPreserveHost On
My plugins all gain a second sam_datasette as in:
http://192.168.0.203/sam_datasette/sam_datasette/rm_toucans_slice.czml?sql=%0A%20%20%20%20%20%20select%0A%20%20rowid%2C%0A%20%20id%2C%0.....
When I modify those instructions and use
ServerName 192.168.0.203:80
ProxyPass /sam_datasette/ http://127.0.0.1:8001/ connectiontimeout=300 timeout=300
ProxyPreserveHost On
this no longer occurs and the plugin points at http://192.168.0.203/sam_datasette/rm_toucans_slice.czml?sql=++++++se...
Consequently, I'm not following the correct httpd.conf instructions.
I'm don't see the same issue on
I've tried with the tables that are already in that database and don't see the issue. Should I, (and can I), try with my tables or template on that database?
Current state for the different versions. Yes/no refer to whether or not the action succeeded.
| Version | table page "Run SQL" | refresh | Query page 'Run SQL' | plugin | ProxyPass |
| 1.0a16 | yes | yes | No. URL not found | yes | Makeshift ProxyPass |
| 1.0a16 | yes | yes | No. URL not found | yes | documented ProxyPass |
On the pre-1.0 version, we get
| Version | table page "Run SQL" | refresh | Query page 'Run SQL' | plugin | ProxyPass |
| 0.65.1 | yes | yes | yes | no: doubles up sam_datasette in the URL as in http://192.168.0.203/sam_datasette//sam_datasette | documented ProxyPass |
| 0.65.1 | yes | yes | yes | yes | MakeshiftProxyPass |
On 1.0a16, remote plugin calls of the form:
rm_toucans_slice.czml?sql=select...
no longer work. Instead, they have to have the form:
rm_toucans_slice/-/query.czml?sql=select...