DWDS/Dialing Concordance Code

+ updated docs for upcoming v2.1.12 file-format changes

+ updated CHitBorders::CBreakCollection::m_BreakOffsets to use mmap()-able ddcVecFile<CTokenNo>, now wrapped with typedef ddcBreakVector

- updated CHitBorders::m_PageBreaks to use mmap()-able ddcVecFile<CPageNumber>

- updated CConcXml::m_Dates to use mmap()-able ddcVecFile<DWORD>

- added CommonLib/ddcStringEnum.h with template class ddcStringEnum<OffT=DWORD> for mmap()-able string<->integer mappings
- updated CConcXml::CFreeBiblStringIndex::m_Values (now m_ValuesE) to use new ddcStringEnum<> with mmap support
+ file format changed: newline-separated "*._bibl_NAME_strings" is now 2 files
~ "*._bibl_NAME_strings_values": NUL-separated values
~ "*._bibl_NAME_strings_idx" : start offsets in *._bibl_NAME_strings_values indexed by logical ID
+ compatibility hack: old "\n"-sparated "*._bibl_NAME_strings" files can still be read in slurp-mode, but produce a warning

+ *perdiv* seems working (still need to test with ddc_index)
+ started work on StringEnum for *._bibl_ATTR_strings (requires conversion, indexing supprt, updated min-compat)

+ added bibl-strings2indexed.perl: upgrade hack for v2.1.12 MemoryMap option

+ added template structs ddcVec2dFile<T>, ddcMapFile<T> in src/CommonLib/ddcMMap.h : for '_perdivToken' & similar ; basic tests seem to work ok (const_iterator is still UNTESTED)

+ mmap: updated CIndexSetForQueryingStage::m_Index (vector<CIndexItem> -> ddcVecFile<CIndexItem>)

+ moved ddcFileMap -> ddcMMap
+ added high-level wrapper template struct ddcVecFile<T> : generic wrapper for slurp- or mmap-vectors