Process large ASCII merger tree files in blocks

[abensonca]

Original report by Andrew Benson (Bitbucket: abensonca, GitHub: abensonca).

---

Using, e.g. Simple_Merger_Tree_File_Maker.exe on very large input files will run out of memory. Could fix this by processing trees in batches.

[abensonca]

Original comment by Andrew Benson (Bitbucket: abensonca, GitHub: abensonca).

---

Another approach (suggested by Jaswant Yadav), would be to allocate space for just one property (e.g. mass, position, etc.) at a time, read that property, then store to file. This might be easier to code, but would require multiple passes through file I/O. It might also be less flexible overall - if you had a really big tree file then it could overflow memory even reading one property at a time, whereas processing in blocks might not.