Problem reading basis set (and question regarding memory)读取基组问题（及内存相关疑问）

Hi, 嗨，

I am having problems reading the basis set from the file ANO-RCC-MB attached. The problem does not occur in an earlier version of Dalton, but with the most recent release that I downloaded today, it gives an error. Output with error is attached.
我在从附带的 ANO-RCC-MB 文件中读取基组时遇到问题。此问题在较早版本的 Dalton 中并未出现，但使用我今天下载的最新版本时，程序会报错。附上包含错误的输出结果。

Now a question about setting the memory (my apologies if this should be put in a separate issue!): What is the limit for setting the memory for a 32-bit integer compiled version of Dalton? I need quite a lot of memory, but Dalton defaults to about 500 MB if I set it too high, e.g. dalton -gb 50.
现在有一个关于设置内存的问题（如果这应该放在单独的问题中，请见谅！）：对于编译为 32 位整数的 Dalton 版本，设置内存的限制是多少？我需要相当多的内存，但如果设置过高，例如 dalton -gb 50，Dalton 默认会限制在约 500 MB。

Best regards Andreas Nyvang serial.out ANO-RCC-MB
此致 安德里亚斯·尼万 serial.out ANO-RCC-MB

changed the description

I can reproduce the error but don't know why it happens. With which version of Dalton do you not get this error?
我可以重现这个错误，但不知道为什么会发生。你使用的是哪个版本的 Dalton 没有出现这个错误？

The memory limit is about 16 GB when using 32-bit integers.
使用 32 位整数时，内存限制约为 16 GB。

The VERSION file says Dalton2020.alpha so I am not entirely sure what version it is. However, in e.g. line 336 in DALTON/abacus/herbas.F I see the following difference: alpha: ANO = .FALSE. ! ANO-RCC is not written in MOLCAS format 2020.0: ANO = .TRUE. So it seems there has been a change in something related to the ANO-RCC basis sets. By the way: the reason that I use ANO-RCC-MB directly instead of contracting the ANO-RCC basis set using the Dalton library is because I haven't been able to get the correct contractions according to the info on ANO basis sets in the chapter on Molecule input format in the manual.
VERSION 文件显示为 Dalton2020.alpha，因此我不完全确定具体是哪个版本。然而，例如在 DALTON/abacus/herbas.F 的第 336 行中，我发现了以下差异：alpha 版本中，ANO = .FALSE. ! ANO-RCC 未以 MOLCAS 格式编写；而 2020.0 版本中，ANO = .TRUE.。由此可见，与 ANO-RCC 基组相关的内容似乎有所改动。顺便提一下：我直接使用 ANO-RCC-MB 而非通过 Dalton 库收缩 ANO-RCC 基组的原因，是因为我未能根据手册中关于分子输入格式章节中 ANO 基组的说明获得正确的收缩结果。

Looks like it was changed by @erikkjellgren. Perhaps @hjaaj can help?
看起来这是由@erikkjellgren 修改的。或许@hjaaj 能帮忙？

I guess the problem is that Dalton expects a particular format for ANO basis sets which your file does not have. A workaround could be to just rename the file though then you should of course check thoroughly that the file is read properly.
我猜问题在于，道尔顿期望 ANO 基组具有特定的格式，而您的文件并不符合这一要求。一个变通的方法可能是直接重命名文件，但当然，您应该彻底检查文件是否被正确读取。

Indeed, renaming the file helps (the format is in Dalton format already from the Basis set Exchange). I also tried using the ANO-RCC-MB in Molcas format (again from Basis set Exchange), but that also seemed not to read in the coefficients correctly.
确实，重命名文件有所帮助（文件格式已从基组交换库中以道尔顿格式提供）。我还尝试使用 Molcas 格式的 ANO-RCC-MB（同样来自基组交换库），但似乎也无法正确读取系数。

I think the Dalton ANO basis format is a Molcas-like format. If you compare https://gitlab.com/dalton/dalton/-/blob/master/basis/ANO-RCC with ANO-RCC Molcas form https://www.basissetexchange.org/, you can see that the basis set exponents have been flattened to a single line, there might also be other small notational differences.
我认为 Dalton ANO 基组格式类似于 Molcas 格式。如果你对比 https://gitlab.com/dalton/dalton/-/blob/master/basis/ANO-RCC 与 ANO-RCC Molcas 格式 https://www.basissetexchange.org/，你会发现基组指数已被压缩至一行，可能还存在其他细微的符号差异。