linear banding strategy

Author: danielplohmann

README.md

@@ -122,6 +122,7 @@ In July 2023, we started populating a [Github repository](https://github.com/dan
 
 ## Version History
 
+ * 2026-01-13 v1.4.6:  Introduced linear banding strategy, to be used as default in future releases (will require re-creation of the whole index).
  * 2025-12-22 v1.4.5:  Fixed a bug due to early conversion when fetching many FunctionEntries at once, which would crash if one function ID does not exist.
  * 2025-12-22 v1.4.4:  No changes, just moved plugins to their own repo located at [mcrit-plugins](https://github.com/danielplohmann/mcrit-plugins).
  * 2025-12-08 v1.4.3:  Major improvements to MCRIT IDA plugin UI, backend now supports faster cross matching jobs only matching among selected samples, minor bugfixes.

mcrit/config/StorageConfig.py

@@ -25,6 +25,10 @@ class StorageConfig(ConfigInterface):
     STORAGE_MONGODB_CLEANUP_TTL: int = 60 * 60 * 24 * 7
     # Once MinHashes have been calculated, discard disassembly from function entries
     STORAGE_DROP_DISASSEMBLY: bool = False
+    # supported strategies: 
+    #  * random: randomly sample from minhash fields, possibly more fuzziness likely won't use all minhash fields
+    #  * linear: use a sequential selection of minhash fields, requires size*number=MINHASH_SIGNATURE_LENGTH
+    STORAGE_BAND_STRATEGY = "random"
     # random seed to be used when deriving sequences used as bands
     STORAGE_BAND_SEED: int = 0xDEADBEEF
     # Banding supports:

mcrit/storage/StorageInterface.py

@@ -686,14 +686,34 @@ def createBandhashProjection(self, minhash):
             a dict containing signature indices used for bandhashing by band id
         """
         band_projection = {}
-        random.seed(self._storage_config.STORAGE_BAND_SEED)
-        band_index = 0
-        for band_size, num_bands in self._storage_config.STORAGE_BANDS.items():
+        banding_strategy = getattr(self._storage_config, "STORAGE_BAND_STRATEGY", "random")
+        if banding_strategy == "random":
+            random.seed(self._storage_config.STORAGE_BAND_SEED)
+            band_index = 0
+            for band_size, num_bands in self._storage_config.STORAGE_BANDS.items():
+                for _ in range(num_bands):
+                    index_sequence = [index for index in range(len(minhash.getMinHashInt()))]
+                    random.shuffle(index_sequence)
+                    band_projection[band_index] = index_sequence[:band_size]
+                    band_index += 1
+        elif banding_strategy == "linear":
+            band_index = 0
+            if len(self._storage_config.STORAGE_BANDS) != 1:
+                raise AttributeError("When using STORAGE_BAND_STRATEGY: linear, use only a single band_size definition - recommended: 4.")
+            size_num_tuples = [i for i in self._storage_config.STORAGE_BANDS.items()][0]
+            band_size = size_num_tuples[0]
+            num_bands = size_num_tuples[1]
+            if not band_size * num_bands == self._minhash_config.MINHASH_SIGNATURE_LENGTH:
+                raise AttributeError("When using STORAGE_BAND_STRATEGY: linear, keep product of band_size (%d) and num_bands (%d) equal to MINHASH_SIGNATURE_LENGTH (%d) - recommended: 4/16/64." % (band_size, num_bands, self._minhash_config.MINHASH_SIGNATURE_LENGTH))
             for _ in range(num_bands):
-                index_sequence = [index for index in range(len(minhash.getMinHashInt()))]
-                random.shuffle(index_sequence)
-                band_projection[band_index] = index_sequence[:band_size]
+                index_sequence = []
+                step_size = int(self._minhash_config.MINHASH_SIGNATURE_LENGTH / band_size)
+                for index_num in range(band_size):
+                    index_sequence.append(index_num * step_size + band_index)
+                band_projection[band_index] = index_sequence
                 band_index += 1
+        else:
+            raise AttributeError("unrecognized STORAGE_BAND_STRATEGY in STORAGE_CONFIG")
         return band_projection
 
     # -> Dict[BandIndex, BandHash]