0xrsydn/idx-cli
1
mirror of https://github.com/0xrsydn/idx-cli.git synced 2026-11-07 01:35:36 +00:00
Code Issues Projects Releases 1 Packages Wiki Activity

fix(core): the IPOT table and date inputs must sit in plain divs

Browse source 
Inside details, dialog, noscript, object, select and the like they may
not show, so the table there rejects the page and such a date input is
no echoed date.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01V3syn2fN2LNo5pvTSqc6eh
This commit is contained in:
hermes 2026-10-18 13:18:30 +00:00
commit 723da4a5a6
1 changed files with 39 additions and 13 deletions
Download patch file Download diff file Expand all files Collapse all files

52
crates/idx-core/src/api/ipot/parse.rs
View file

@ -163,10 +163,21 @@ fn cell(td: ElementRef) -> Option<(String, Option<String>)> {
Some((clean(&text), class))
}
/// The `MM/DD/YYYY` value of the one input with this id.
/// Whether `element` sits in plain divs only, as IPOT's table and date
/// inputs do; inside anything else (details, dialog, noscript, object,
/// select, ...) it may not show.
fn in_divs(element: ElementRef) -> bool {
element
.ancestors()
.filter_map(ElementRef::wrap)
.all(|ancestor| matches!(ancestor.value().name(), "div" | "html"))
}
/// The `MM/DD/YYYY` value of the one input with this id, in plain divs.
fn input_date(page: &Html, id: &str) -> Option<NaiveDate> {
let value = only(page.root_element(), &format!("input#{id}"))?.attr("value")?;
NaiveDate::parse_from_str(value.trim(), "%m/%d/%Y").ok()
let input =
only(page.root_element(), &format!("input#{id}")).filter(|input| in_divs(*input))?;
NaiveDate::parse_from_str(input.attr("value")?.trim(), "%m/%d/%Y").ok()
}
/// Whether anything in the page is hidden by markup alone (`hidden`, an
@ -200,6 +211,9 @@ pub fn parse_broker_summary(html: &str) -> Result<IpotPage, IdxError> {
}
let table = only(page.root_element(), "table")
.ok_or_else(|| parse_error("no single table in the page"))?;
if !in_divs(table) {
return Err(parse_error("the table is not in plain page markup"));
}
let body = only(table, "tbody").ok_or_else(|| parse_error("no single table body"))?;
let mut raw = BrokerSummaryRaw {
from: input_date(&page, "startDateBrokerSummary"),
@ -490,6 +504,22 @@ mod tests {
1,
),
),
(
"the table in a select",
DAY.replacen("<table", "<select><table", 1).replacen(
"</table>",
"</table></select>",
1,
),
),
(
"the table in closed details",
DAY.replacen("<table", "<details><table", 1).replacen(
"</table>",
"</table></details>",
1,
),
),
("no table", "</tbody><tbody>".to_string()),
// Text a browser would not show, or not as one figure or code.
(
@ -589,16 +619,6 @@ mod tests {
1,
),
),
// The HTML standard keeps a table inside a select since its 2025
// select parsing change; html5ever and current browsers agree.
(
"the table in a select",
DAY.replacen("<table", "<select><table", 1).replacen(
"</table>",
"</table></select>",
1,
),
),
(
"a script after the table",
DAY.replacen("</tbody>", "</tbody><script></script>", 1),
@ -626,5 +646,11 @@ mod tests {
None,
"an id twice is no echoed date"
);
let closed = DAY.replacen("<input", "<details><input", 1);
assert_eq!(
summary(&closed).from,
None,
"a date in closed details is not shown"
);
}
}