fix: qe input blocks not seperated by empty lines (#724)

the qe support only parse input file, in which cell blocks are separated
by empty lines, like
```
ATOMIC_SPECIES
 Na  22.989769  Na_ONCV_PBE-1.0.upf
                                                          
CELL_PARAMETERS {angstrom}
 7.171683039200000  0.000000000000000  0.000000000000000    
```
however, the input file is valid when no empty line exists, like the
following
```
ATOMIC_SPECIES
 Na  22.989769  Na_ONCV_PBE-1.0.upf  
CELL_PARAMETERS {angstrom} 
 7.171683039200000  0.000000000000000  0.000000000000000    
```
This pr fixes the issue

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

## Summary by CodeRabbit

- **New Features**
- Improved handling of Quantum Espresso output data for better parsing
and clarity.
	- Enhanced error handling for missing force and stress data.
- Added configuration parameters for self-consistent field (SCF)
calculations for sodium.

- **Bug Fixes**
- Updated functions to prevent unwanted lines in data extraction,
ensuring cleaner output.

- **Tests**
- Introduced a new test class to validate the functionality of the
`dpdata.LabeledSystem` class with Quantum Espresso output, enhancing
test coverage.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->

---------

Co-authored-by: Han Wang <[email protected]>

Author: wanghan-iapcm

dpdata/qe/scf.py

@@ -11,15 +11,32 @@
 bohr2ang = 0.52917721067
 kbar2evperang3 = 1e3 / 1.602176621e6
 
+_QE_BLOCK_KEYWORDS = [
+    "ATOMIC_SPECIES",
+    "ATOMIC_POSITIONS",
+    "K_POINTS",
+    "ADDITIONAL_K_POINTS",
+    "CELL_PARAMETERS",
+    "CONSTRAINTS",
+    "OCCUPATIONS",
+    "ATOMIC_VELOCITIES",
+    "ATOMIC_FORCES",
+    "SOLVENTS",
+    "HUBBARD",
+]
+
 
 def get_block(lines, keyword, skip=0):
     ret = []
     for idx, ii in enumerate(lines):
         if keyword in ii:
             blk_idx = idx + 1 + skip
-            while len(lines[blk_idx]) == 0:
+            while len(lines[blk_idx].split()) == 0:
                 blk_idx += 1
-            while len(lines[blk_idx]) != 0 and blk_idx != len(lines):
+            while (
+                len(lines[blk_idx].split()) != 0
+                and (lines[blk_idx].split()[0] not in _QE_BLOCK_KEYWORDS)
+            ) and blk_idx != len(lines):
                 ret.append(lines[blk_idx])
                 blk_idx += 1
             break
@@ -123,7 +140,7 @@ def get_force(lines, natoms):
 def get_stress(lines):
     blk = get_block(lines, "total   stress")
     if len(blk) == 0:
-        return
+        return None
     ret = []
     for ii in blk:
         ret.append([float(jj) for jj in ii.split()[3:6]])

tests/qe.scf/na.in

@@ -0,0 +1,44 @@
+
+&control
+    calculation = 'scf'
+    restart_mode = 'from_scratch',
+    prefix = 'Na'
+    outdir = './tmp'
+    pseudo_dir = './'
+/
+&SYSTEM
+ ibrav=0,   
+ nosym=.true.,                     
+ nat=3,                         
+ ntyp=1,                        
+ occupations = 'smearing',       
+ smearing = 'gauss',                
+ degauss = 0.02,                   
+ ecutwfc = 100,                   
+ ecutrho = 960,                   
+ lspinorb = .false.,                
+ noncolin = .false.,                
+/
+&ELECTRONS
+ conv_thr = 1.0d-9,                  
+ mixing_beta = 0.7,               
+ electron_maxstep = 200,          
+/
+&ions
+/
+&cell
+/
+ATOMIC_SPECIES
+ Na  22.989769  Na_ONCV_PBE-1.0.upf
+CELL_PARAMETERS {angstrom}        
+ 7.171683039200000  0.000000000000000  0.000000000000000
+ -4.270578118300000  5.761527588200000  0.000000000000000
+ -0.000000045600000  0.000000023000000  12.826457854999999
+ATOMIC_POSITIONS (crystal)       
+   
+ Na  0.940587444301534  0.397635863676890  0.059472381901808
+ Na  0.059413515648061  0.602364552456546  0.559472465518034
+ Na  0.602364619812068  0.059413062489401  0.059472381901808
+K_POINTS {automatic} 
+ 8 8 4  0 0 0
+