-
Taking a CML file of 5.5GB (corresponding to approx 2 million) structures crashes the stdinchi-1 v1.2 linux application. It throws a std::bad_alloc, i.e., runs out of memory. With smaller files it's alright. It seems stdinchi-1 tries to read the whole file into memory. I haven't checked in the source. But given that it runs with smaller files, this doesn't sound too unreasonable.
2009-04-24 13:16:42 UTC by nobody
-
uLUYft rigriemdyoej, [url=http://hnvmcclbzqkd.com/]hnvmcclbzqkd[/url], [link=http://ikwkbpzqnbzi.com/]ikwkbpzqnbzi[/link], http://ufcywphzllqn.com/.
2008-11-28 03:01:22 UTC by nobody
-
(for version 1.01 on Windows Vista)
Calling the cInChI-1.exe with the /InChI2Struct option from within a Java program (see example below) seems to "hang" while waiting for the process to finish, despite the fact that all the files have been written and the output indicates that processing is finished. Maybe this is a bug and the process is not returning an appropriate exit value with the...
2008-09-05 17:38:42 UTC by gmagoon
-
I need to be able to copy/ paste the InChIKey into an Excel spreadsheet. With version 1.01 /Tabbed got
me the output I needed, but it doesn't work with InChIKey. The InChIKey appears on a separate line beneath the InChI and AuxInfo.
A sample output is attached.
2007-11-06 14:37:52 UTC by maritja
-
From the mailing list:
---------------------------- Original Message ----------------------------
Subject: Re: [InChI-discuss] Packaging inchi - licensing
issues clarification request
Date: Mon, August 13, 2007 8:35 pm
To: inchi-discuss@lists.sourceforge.net
--------------------------------------------------------------------------
Alan,
On Friday, 10 August 2007 at...
2007-08-17 15:25:39 UTC by egonw
-
I noticed an elegancy bug in ichiparm.h:DetectInputINChIFileType. In addition to the inelgancy of having non-inline function implementation in a header file, there's code which looks like this:
static char szPlnVersion[NUM_VERSIONS][LEN_VERSIONS]; /* = "INChI:1.1Beta/";
...
static int bInitilized = 0;
...
if ( !bInitilized ) {
lenPlnVersion[0] = sprintf(...
2007-07-07 00:12:48 UTC by dalke
-
I agree, it's not a bug. This is a feature request.
Given that the input is not an InChI it is of course outside the spec to determine how to report the error. The code could as easily say that the connectivity is wrong, or the charges.
As I worked to understand InChI strings, by generating them by hand or through my own software assist, I found the message "Syntax error (-2) in...
2007-07-02 21:37:26 UTC by dalke
-
Mobile H formula means chemical formula of the mobile H layer, in this case "Q".
Although one may call this somewhat cryptic, in my opinion it is not a bug.
- Dmitrii.
2007-07-02 20:25:21 UTC by nobody
-
When using the code to convert InChI2InChI I found that
InChI=1/CH4C/c1-2/h1H2,2H2
remains as
InChI=1/CH4C/c1-2/h1H2,2H2
when I expected it to be
InChI=1/C2H4/c1-2/h1-2H2
and I found that
InChI=1/CO/c1-2
InChI=1/OC/c1-2
do not produce identical output. Indeed, they stay unchanged.
I thought that the -InChI2InChI would perform a sort of canonicalization of InChI-like...
2007-07-01 04:39:59 UTC by dalke
-
The internal charge parser code uses 9999 to mean that no charge was specified. So the totally unchemical structure InChI=1/CO/q+9999 produces InChI=1/CO
This probably isn't a concern except in contrived circumstances.
2007-07-01 02:32:34 UTC by dalke
