Re: Why does the prettyprint of std::tuple start with [1]?
David Blaikie via Gdb <[email protected]>
| Newsgroups | gmane.comp.gdb.devel |
|---|---|
| Message-ID | <CAENS6Es6NX-2WJbNBSEj8UG1OYEsvG9ueRxQw7KxUnA6qtZkQA@mail.gmail.com> |
Likely it does, yeah. Probably a bug in libstdc++'s pretty printer: https://github.com/gcc-mirror/gcc/blob/8a7d54b1e10b8f4fba1358260ed2e7056ed23cbd/libstdc%2B%2B-v3/python/libstdcxx/v6/printers.py#L528 https://github.com/gcc-mirror/gcc/blob/8a7d54b1e10b8f4fba1358260ed2e7056ed23cbd/libstdc%2B%2B-v3/python/libstdcxx/v6/printers.py#L563 Something related to those bits of code. On Sun, Jun 13, 2021 at 8:37 AM Paul Smith <[email protected]> wrote: > On Sun, 2021-06-13 at 09:49 -0400, Paul Smith wrote: > > Am I the only one to be confused by the fact that the pretty-printer > > for std::tuple values starts indexing at 1 instead of 0? > > > > (gdb) p info > > $1 = (const std::tuple<TestType, DataType> &) @0x7fffffffd550: { > > param = std::tuple containing = { > > [1] = TestType::OneTest, > > [2] = DataType::SomeData > > }, > > index = 2 > > } > > > > Why aren't the indexes on the left numbered [0] and [1], > > corresponding > > to the values used with std::get(), instead of [1] and [2]? > > Hm, maybe this belongs on the GCC mailing list instead of here. > >